Active Vibration Control of a Modular Robot Combining a Back-Propagation Neural Network with
Li, Yangmin
by joints, vibration can easily be induced in this special type of mechanical structure. Based on the modalActive Vibration Control of a Modular Robot Combining a Back-Propagation Neural Network-propagation neural network suboptimal controller is developed to control the vibration of a nine
Handwritten Digit Recognition with a Back-Propagation Network
Parker, Gary B.
Handwritten Digit Recognition with a Back-Propagation Network Y. Le Cun, B. Boser, J. S. Denker, D We present an application of back-propagation networks to hand- written digit recognition. Minimal. 1 INTRODUCTION The main point of this paper is to show that large back-propagation (BP) net- works
Handwritten Digit Recognition with a BackPropagation Network
LeCun, Yann
Handwritten Digit Recognition with a BackÂPropagation Network Y. Le Cun, B. Boser, J. S. Denker, D We present an application of backÂpropagation networks to handÂ written digit recognition. Minimal. 1 INTRODUCTION The main point of this paper is to show that large backÂpropagation (BP) netÂ works
Contextual Back-Propagation Technical Report UT-CS-00-443
MacLennan, Bruce
Contextual Back-Propagation Technical Report UT-CS-00-443 Bruce J. MacLennan #3; Computer Science. This report presents an adaptation of the back- propagation algorithm to training contextual neural networks and adaptation must also be context- dependent. The basic idea is simple enough | hold the context constant while
Neural Networks, Vol. 10, NO. 8, 1505-1521, 1997pp. 0 1997 Elsevier Science Ltd. All rights reserved
Edinburgh, University of
Pergamon Neural Networks, Vol. 10, NO. 8, 1505-1521, 1997pp. 0 1997 Elsevier Science Ltd. All learning parameters on the learning procedure and pe$ormance of back-propagation neural networks used Elsevier Science Ltd. Keywords-Back-propagation neural network, Seismic arrivals picking, Pattern
An Empirical Study of Learning Speed in BackPropagation Networks
Fahlman, Scott E.
the basic ideas of connectionism or backÂpropagation learning. See [3] for a brief overview of this areaAn Empirical Study of Learning Speed in BackÂPropagation Networks Scott E. Fahlman September 1988 of the backÂpropagation algorithm. However, backÂpropagation learning is too slow for many applications
Development of a potential field estimator for a path-planning application using neural networks
Smith, Darin William
1997-01-01
This thesis presents the development of a potential field estimator for a locally constrained autonomous path-planning application. The potential field estimator was developed using back-propagation neural networks, which ...
Image Compression by Back Propagation
Cottrell, Garrison W.
CHAPTER 9 Image Compression by Back Propagation: An Example of Extensional Programming* GARRISON W the case with the computatiolls associated with basic cognitive pro- cesses such as vision and audition techniques. The technique we employ is known as back propagation. developed by l1umelhart, Hinton
Computationally Efficient Neural Network Intrusion Security Awareness
Todd Vollmer; Milos Manic
2009-08-01
An enhanced version of an algorithm to provide anomaly based intrusion detection alerts for cyber security state awareness is detailed. A unique aspect is the training of an error back-propagation neural network with intrusion detection rule features to provide a recognition basis. Network packet details are subsequently provided to the trained network to produce a classification. This leverages rule knowledge sets to produce classifications for anomaly based systems. Several test cases executed on ICMP protocol revealed a 60% identification rate of true positives. This rate matched the previous work, but 70% less memory was used and the run time was reduced to less than 1 second from 37 seconds.
Neural Networks Early Neural Network Modeling
Yuste, Rafael
Appendix E Neural Networks Early Neural Network Modeling Neurons Are Computational Devices A Neuron? This is the central question moti- vating the study of neural networks. In this appendix we provide a brief historical review of the field, intro- duce some key concepts, and discuss two influential models of neural networks
Gabel, S.
2003-01-01
A neural network approach is employed for estimating key efficiency parameters in a gas turbine engine. The concept is demonstrated within a limited operating region for a given engine. The neural network is developed to estimate certain...
Analysis of neutron noise spectra using neural networks
Korsah, K. (Oak Ridge National Lab., TN (USA)); Uhrig, R.E. (Oak Ridge National Lab., TN (USA) Tennessee Univ., Knoxville, TN (USA))
1991-01-01
Neural network architectures based on the back-propagation paradigm have been developed to recognize the features, and detect resonance shifts in, power spectral density (PSD) data. Our goal is to advance the state of the art in the application of noise analysis techniques to monitor nuclear reactor internals. The initial objectives have been to use PSD data, acquired over a period of about 2 years by PSDREC (power spectral density recognition system), to develop neural networks that are able to differentiate between normal neutron power spectral density data and anomalous spectral data, and detect significant shifts in the positions of spectral resonances while reducing the effect of small shifts. Neural network systems referred to in this paper as spectral feature detectors (SFDs) and integral network filters have been developed to meet these objectives. The performance of the SFDs is the subject of this paper. 2 refs., 2 figs.
Discrimination of neutrons and {\\gamma}-rays in liquid scintillator based on Elman neural network
Zhang, Cai-Xun; Zhao, Jian-Ling; Wang, Li; Yu, Xun-Zhen; Zhu, Jing-Jun; Xing, Hao-Yang
2015-01-01
A new neutron and {\\gamma} (n/{\\gamma}) discrimination method based on Elman Neural Network (ENN) was put forward to improve the n/{\\gamma} discrimination performance of liquid scintillator (LS). In this study, neutron and {\\gamma} data acquired from EJ-335 which was exposed in Am-Be radiation field was discriminated using ENN. The difference of n/{\\gamma} discrimination performance between using ENN and Back Propagation Neural Network (BPNN) is that ENN gave a improvement over BPNN in n/{\\gamma} discrimination with the increasing increasing of the Figure of Merit (FOM) from 0.907 to 0.953.
Misbehavior in a Neural Network Model
Burgos, José E
2015-01-01
Rodríguez, E. (2007). Neural-network simulations of twodiscipline of will. Neural Networks, 19, 1153–1160. http://E. R. Staddon (Eds. ), Neural network models of conditioning
Solos (Dice Game) and Conductor (Neural Network)
Marquetti, Andre
2015-01-01
vice versa). Figure 6.1 Neural Network keyboard input styleamplitude, pitch-class, neural network key patterns dependconvergence with the neural network does not depend on the
Radu Calin Dimitriu NEURAL NETWORKS,
Cambridge, University of
Radu Calin Dimitriu NEURAL NETWORKS, TRICKS OF THE TRADE #12;BEFORE STARTING THE TRAINING A DATABASE FOR A NEURAL NETWORK THIS WILL ENSURE SUFFICIENT DATA AND VARIABLES TO CAPTURE THE COMPLEXITY IN THE DATABASE IF YOU DO NOT KNOW HOW THEY ARE RELATED TO THE OUTPUT, THE NEURAL NETWORK MAY NEVERTHELESS
Symbolic Representation of Neural Networks
Liu, Huan
Draft Symbolic Representation of Neural Networks Rudy Setiono and Huan Liu Department,liuhg@iscs.nus.sg Abstract Although backpropagation neural networks generally predict better than decision trees do is needed by human experts. This work drives a sym bolic representation for neural networks to make
Neural Network Based Intrusion Detection System for Critical Infrastructures
Todd Vollmer; Ondrej Linda; Milos Manic
2009-07-01
Resiliency and security in control systems such as SCADA and Nuclear plant’s in today’s world of hackers and malware are a relevant concern. Computer systems used within critical infrastructures to control physical functions are not immune to the threat of cyber attacks and may be potentially vulnerable. Tailoring an intrusion detection system to the specifics of critical infrastructures can significantly improve the security of such systems. The IDS-NNM – Intrusion Detection System using Neural Network based Modeling, is presented in this paper. The main contributions of this work are: 1) the use and analyses of real network data (data recorded from an existing critical infrastructure); 2) the development of a specific window based feature extraction technique; 3) the construction of training dataset using randomly generated intrusion vectors; 4) the use of a combination of two neural network learning algorithms – the Error-Back Propagation and Levenberg-Marquardt, for normal behavior modeling. The presented algorithm was evaluated on previously unseen network data. The IDS-NNM algorithm proved to be capable of capturing all intrusion attempts presented in the network communication while not generating any false alerts.
Study of a transient identification system using a neural network for a PWR plant
Ishihara, Yoshinao; Kasai, Masao; Kambara, Masayuki [Mitsubishi Heavy Industries, Ltd., Yokohama (Japan); Mitsuda, Hiromichi; Kurata, Toshikazu; Shirosaki, Hidekazu [Inst. of Nuclear Safety System, Inc., Kyoto (Japan)
1996-08-01
This paper presents the procedure and results of a system for identifying PWR plant abnormal events, which uses neural network techniques. The neural network recognizes the abnormal event from the patterns of the transient changes of analog data from plant parameters when they deport from their normal state. For the identification of abnormal events in this study, events that cause a reactor to scram during power operation were selected as the design base events. The test data were prepared by simulating the transients on a compact PWR simulator. The simulation data were analyzed to determine how the plant parameters respond after the occurrence of a transient. A method of converting the pattern of the transient changes into characteristic parameters by fitting the data to pre-determined functions was developed. These characteristic parameters were used as the input data to the neural network. The neural network learning procedure used a generalized delta rule, namely a back-propagation algorithm. The neural network can identify the type of an abnormal event from a limited set of events by using these characteristic parameters obtained from the pattern of the changes in the analog data. From the results of this application of a neural network, it was concluded that it would be possible to use the method to identify abnormal events in a nuclear power plant.
Neural Pascal A Language for Neural Network Programming
Gumm, H. Peter
Neural Pascal A Language for Neural Network Programming H.Peter Gumm Dept. of Computer Science SUNY is an extension of object-oriented Pascal, designed to allow easy specification and simulation of neural networks through all elements of a linear datatype. 1. Introduction Neural network models are frequently visualized
Introduction to Artificial Intelligence Neural Networks
Qu, Rong
Introduction to Artificial Intelligence (G51IAI) Dr Rong Qu Neural Networks #12;G51IAI Introduction to AI Neural Networks Chapter 20 Artificial Intelligence : A Modern Approach (AIMA) Russell Introduction to AI Neural Networks More precisely: Artificial Neural Networks Simulating, on a computer, what
Artificial Neural Network for Optimized Power System Management
OLeary, Daniel Albert
2015-01-01
vii Abstract Artificial Neural Network for Optimized PowerAn Artificial Neural Network (ANN). Data is input to theSANTA CRUZ ARTIFICIAL NEURAL NETWORK FOR OPTIMIZED POWER
An efficient online feature extraction algorithm for neural networks
Bozorgmehr, Pouya
2009-01-01
2.3 Artificial Neural Networks . . . . . . . . . . .2.3.1 Learning in Artificial Neural Networks 2.3.2 HopfieldBayesian learning for neural networks. Springer Verlag,
Keeni, Kanad; Nakayama, Kenji; Shimodaira, Hiroshi
This study high lights on the subject of weight initialization in back-propagation feed-forward networks. Training data is analyzed and the notion of critical points is introduced for determining the initial weights and ...
Mjolsness, Eric
Symbolic Neural Networks Derived from Stochastic Grammar Domain Models 1 Symbolic Neural Networks neural network architectures with some of the expressive power of a semantic network and also some of the pattern recognition and learning capabilities of more conventional neural networks. For example
Comparison between Traditional Neural Networks and Radial Basis Function Networks
Wilamowski, Bogdan Maciej
Comparison between Traditional Neural Networks and Radial Basis Function Networks Tiantian Xie, Hao networks: traditional neural networks and radial basis function (RBF) networks, both of which of neural network architectures are analyzed and compared based on four different examples. The comparison
Backpropagation in Sequential Deep Neural Networks
Noble, William Stafford
Backpropagation in Sequential Deep Neural Networks Galen Andrew University of Washington galen neural networks to problems in speech processing has combined the output of a static network trained over developed Sequential Deep Neural Network (SDNN) model allows sequential dependencies between internal hidden
Aircraft System Identification Using Artificial Neural Networks
Valasek, John
Aircraft System Identification Using Artificial Neural Networks Kenton Kirkpatrick Jim May Jr. John Networks 2 Artificial Neural Networks ANNSID Conclusions and Open Challenges #12;Motivation 3 #12;Motivating Questions Is it possible to use artificial neural networks to determine a linear model
Designing Neural Networks Using Gene Expression Programming
Fernandez, Thomas
1 Designing Neural Networks Using Gene Expression Programming CândidaFerreira Gepsoft, 73 Elmtree aspects of neural networks, such as the weights, the thresholds, and the network architec- ture. Indeed neural network, including the architecture, the weights and thresholds, could be totally encoded
Connectivity in the Yeast Cell Cycle Transcription Network: Inferences from Neural Networks
Hart, Christopher E; Mjolsness, Eric; Wold, Barbara J
2006-01-01
Issue 12 | e169 Neural Network Model of Yeast Transcription27. Bishop C (1995) Neural networks for pattern recognition.Network: Inferences from Neural Networks Christopher E. Hart
Advanced battery modeling using neural networks
Arikara, Muralidharan Pushpakam
1993-01-01
battery models are available today that can accurately predict the performance of the battery system. This thesis presents a modeling technique for batteries employing neural networks. The advantage of using neural networks is that the effect of any...
Solos (Dice Game) and Conductor (Neural Network)
Marquetti, Andre
2015-01-01
method to compose music and the Conductors’ neural network for processing the music is based on pattern- extraction,
Quantum neural networks Alexandr A. Ezhov1
Martinez, Tony R.
Quantum neural networks Alexandr A. Ezhov1 and Dan Ventura2 1 Department of Mathematics, Troitsk outlines the research, development and perspectives of quantum neural networks a burgeoning new field of quantum neural networks may give us both new undestanding of brain function as well as unprecedented
Consistency of Posterior Distributions for Neural Networks
Consistency of Posterior Distributions for Neural Networks Herbert Lee \\Lambda May 21, 1998 Abstract In this paper we show that the posterior distribution for feedforward neural networks is asymp neural networks for nonparametric regression in a Bayesian framework. Keywords: Bayesian statistics
Supervised Learning in Neural Networks without Feedback Networks
Lin, Feng
Supervised Learning in Neural Networks without Feedback Networks Robert D. Brandt and Feng Lin Abstract In this paper, we study the supervised learning in neural networks. Unlike the com- mon practice (hardware) implementation of arti cial neural networks. This research is supported in part by the National
Christo, F.C.; Masri, A.R.; Nebot, E.M.
1996-09-01
A novel approach using artificial neural networks for representing chemical reactions is developed and successfully implemented with a modeled velocity-scalar joint pdf transport equation for H{sub 2}CO{sub 2} turbulent jet diffusion flames. The chemical kinetics are represented using a three-step reduced mechanism, and the transport equation is solved by a Monte Carlo method. A detailed analysis of computational performance and a comparison between the neural network approach and other methods used to represent the chemistry, namely the look-up table, or the direct integration procedures, are presented. A multilayer perceptron architecture is chosen for the neural network. The training algorithm is based on a back-propagation supervised learning procedure with individual momentum terms and adaptive learning rate adjustment for the weights matrix. A new procedure for the selection of training samples using dynamic randomization is developed and is aimed at reducing the possibility of the network being trapped in a local minimum. This algorithm achieved an impressive acceleration in convergence compared with the use of a fixed set of selected training samples. The optimization process of the neural network is discussed in detail. The feasibility of using neural network models to represent highly nonlinear chemical reactions is successfully illustrated. The prediction of the flow field and flame characteristics using the neural network approach is in good agreement with those obtained using other methods, and is also in reasonable agreement with the experimental data. The computational benefits of the neural network approach over the look-up table and the direct integration methods, both in CPU time and RAM storage requirements are not great for a chemical mechanisms of less than three reactions. The neural network approach becomes superior, however, for more complex reaction schemes.
Nonlinear principal component analysis by neural networks
Hsieh, William
Nonlinear principal component analysis by neural networks William W. Hsieh Oceanography oversimplification of the datasets being analyzed. The advent of neural network (NN) models, a class of powerful by a neural net- work model which nonlinearly generalizes the classical principal component analysis (PCA
Nonlinear principal component analysis by neural networks
Hsieh, William
Nonlinear principal component analysis by neural networks William W. Hsieh Oceanography a potential oversimplification of the datasets being analyzed. The advent of neural network (NN) models by a neural net work model which nonlinearly generalizes the classical principal component analysis (PCA
Identification and control of plasma vertical position using neural network in Damavand tokamak
Rasouli, H. [School of Plasma Physics and Nuclear Fusion, Institute of Nuclear Science and Technology, AEOI, P.O. Box 14155-1339, Tehran (Iran, Islamic Republic of); Advanced Process Automation and Control (APAC) Research Group, Faculty of Electrical Engineering, K.N. Toosi University of Technology, P.O. Box 16315-1355, Tehran (Iran, Islamic Republic of); Rasouli, C.; Koohi, A. [School of Plasma Physics and Nuclear Fusion, Institute of Nuclear Science and Technology, AEOI, P.O. Box 14155-1339, Tehran (Iran, Islamic Republic of)
2013-02-15
In this work, a nonlinear model is introduced to determine the vertical position of the plasma column in Damavand tokamak. Using this model as a simulator, a nonlinear neural network controller has been designed. In the first stage, the electronic drive and sensory circuits of Damavand tokamak are modified. These circuits can control the vertical position of the plasma column inside the vacuum vessel. Since the vertical position of plasma is an unstable parameter, a direct closed loop system identification algorithm is performed. In the second stage, a nonlinear model is identified for plasma vertical position, based on the multilayer perceptron (MLP) neural network (NN) structure. Estimation of simulator parameters has been performed by back-propagation error algorithm using Levenberg-Marquardt gradient descent optimization technique. The model is verified through simulation of the whole closed loop system using both simulator and actual plant in similar conditions. As the final stage, a MLP neural network controller is designed for simulator model. In the last step, online training is performed to tune the controller parameters. Simulation results justify using of the NN controller for the actual plant.
Prediction of Weld Penetration in FCAW of HSLA steel using Artificial Neural Networks
Asl, Y. Dadgar; Mostafa, N. B.; Panahizadeh, V. R. [Department of Mechanical Engineering, Shahid Rajaee Teacher Training University, Tehran (Iran, Islamic Republic of); Seyedkashi, S. M. H. [Department of Mechanical Engineering, Tarbiat Modares University, Tehran (Iran, Islamic Republic of)
2011-01-17
Flux-cored arc welding (FCAW) is a semiautomatic or automatic arc welding process that requires a continuously-fed consumable tubular electrode containing a flux. The main FCAW process parameters affecting the depth of penetration are welding current, arc voltage, nozzle-to-work distance, torch angle and welding speed. Shallow depth of penetration may contribute to failure of a welded structure since penetration determines the stress-carrying capacity of a welded joint. To avoid such occurrences; the welding process parameters influencing the weld penetration must be properly selected to obtain an acceptable weld penetration and hence a high quality joint. Artificial neural networks (ANN), also called neural networks (NN), are computational models used to express complex non-linear relationships between input and output data. In this paper, artificial neural network (ANN) method is used to predict the effects of welding current, arc voltage, nozzle-to-work distance, torch angle and welding speed on weld penetration depth in gas shielded FCAW of a grade of high strength low alloy steel. 32 experimental runs were carried out using the bead-on-plate welding technique. Weld penetrations were measured and on the basis of these 32 sets of experimental data, a feed-forward back-propagation neural network was created. 28 sets of the experiments were used as the training data and the remaining 4 sets were used for the testing phase of the network. The ANN has one hidden layer with eight neurons and is trained after 840 iterations. The comparison between the experimental results and ANN results showed that the trained network could predict the effects of the FCAW process parameters on weld penetration adequately.
Master Thesis Combined Neural Networks and
Cambridge, University of
Master Thesis Combined Neural Networks and Genetic Algorithms as a method for reducing redundancyNetworksandGentworksandGentworksandGentworksandGeneticAlgorithmseticAlgorithmseticAlgorithmseticAlgorithms asaasaasaasamethodformethodformethodformethodforreducingredundancyinsteeldesignreducingredundancyinsteeldesignreducingredundancyinsteeldesignreducingredundancyinsteeldesign 2008MinSungJoo2008MinSungJoo2008MinSungJoo2008MinSungJoo #12; Combined Neural Networks and Genetic Algorithms as a method for reducing redundancy in steel design #12;Combined Neural Networks
Neural network based system for equipment surveillance
Vilim, Richard B. (Aurora, IL); Gross, Kenneth C. (Bolingbrook, IL); Wegerich, Stephan W. (Glendale Hts., IL)
1998-01-01
A method and system for performing surveillance of transient signals of an industrial device to ascertain the operating state. The method and system involves the steps of reading into a memory training data, determining neural network weighting values until achieving target outputs close to the neural network output. If the target outputs are inadequate, wavelet parameters are determined to yield neural network outputs close to the desired set of target outputs and then providing signals characteristic of an industrial process and comparing the neural network output to the industrial process signals to evaluate the operating state of the industrial process.
Neural network based system for equipment surveillance
Vilim, R.B.; Gross, K.C.; Wegerich, S.W.
1998-04-28
A method and system are disclosed for performing surveillance of transient signals of an industrial device to ascertain the operating state. The method and system involves the steps of reading into a memory training data, determining neural network weighting values until achieving target outputs close to the neural network output. If the target outputs are inadequate, wavelet parameters are determined to yield neural network outputs close to the desired set of target outputs and then providing signals characteristic of an industrial process and comparing the neural network output to the industrial process signals to evaluate the operating state of the industrial process. 33 figs.
Characterization of Shape Memory Alloys Using Artificial Neural Networks
Valasek, John
1 Characterization of Shape Memory Alloys Using Artificial Neural Networks Jim Henrickson, Kenton Generate Training Data Train Artificial Neural Network Results Conclusion Characterization of Shape Characterization of Shape Memory Alloys Using Artificial Neural Networks Jim Henrickson, Kenton Kirkpatrick, Dr
USING NEURAL NETWORKS FOR WEB PROXY CACHE REPLACEMENT
ElAarag, Hala
USING NEURAL NETWORKS FOR WEB PROXY CACHE REPLACEMENT by JAKE COBB Advisor HALA ELAARAG A senior................................................................................................... 10 4. NEURAL NETWORK PROXY CACHE REPLACEMENT .................................. 12 5. SIMULATION replacement is developed. Unlike previous approaches, this research utilizes neural networks for replacement
CNNcon: Improved Protein Contact Maps Prediction Using Cascaded Neural Networks
2013-01-01
Top L/5 predicted. PROFcon j Neural network based; Top L/2of proteins with neural networks and correlated mutations.using 2D-recursive neural networks. Nucleic Acids Research
Optimization and global minimization methods suitable for neural networks
Neumaier, Arnold
Optimization and global minimization methods suitable for neural networks Wlodzislaw Duch and Jerzy Louis Pasteur, Blvd. Sebastien Brant, 67400 Illkirch, France Abstract Neural networks are usually and statistical methods in many applications. For neural networks with predetermined structure, for example
Neural Networks for Adaptive Processing of Structured Data
Sperduti, Alessandro
Neural Networks for Adaptive Processing of Structured Data Alessandro Sperduti Dip. di Informatica Recursive Neural Networks, i.e. neural network models able to deal with data represented as directed acyclic capabilities of neural networks to structured domains. While earlier neural approaches were able to deal
Artificial neural networks in models of specialization, guild evolution
Getz, Wayne M.
Artificial neural networks in models of specialization, guild evolution and sympatric speciation on host choice, employing artificial neural networks as models for the host recognition system
Neural network approach to parton distributions fitting
Andrea Piccione; Joan Rojo; for the NNPDF Collaboration
2005-10-18
We will show an application of neural networks to extract information on the structure of hadrons. A Monte Carlo over experimental data is performed to correctly reproduce data errors and correlations. A neural network is then trained on each Monte Carlo replica via a genetic algorithm. Results on the proton and deuteron structure functions, and on the nonsinglet parton distribution will be shown.
Enhancing neural-network performance via assortativity
Franciscis, Sebastiano de; Johnson, Samuel; Torres, Joaquin J. [Departamento de Electromagnetismo y Fisica de la Materia, and Institute Carlos I for Theoretical and Computational Physics, and Facultad de Ciencias, University of Granada, E-18071 Granada (Spain)
2011-03-15
The performance of attractor neural networks has been shown to depend crucially on the heterogeneity of the underlying topology. We take this analysis a step further by examining the effect of degree-degree correlations - assortativity - on neural-network behavior. We make use of a method recently put forward for studying correlated networks and dynamics thereon, both analytically and computationally, which is independent of how the topology may have evolved. We show how the robustness to noise is greatly enhanced in assortative (positively correlated) neural networks, especially if it is the hub neurons that store the information.
VLSI Cells Placement Using the Neural Networks
Azizi, Hacene; Zouaoui, Lamri; Mokhnache, Salah [Universite Ferhat Abbas, Faculte des Sciences Laboratoire Optoelectronique et Composants, Setif(Algeria)
2008-06-12
The artificial neural networks have been studied for several years. Their effectiveness makes it possible to expect high performances. The privileged fields of these techniques remain the recognition and classification. Various applications of optimization are also studied under the angle of the artificial neural networks. They make it possible to apply distributed heuristic algorithms. In this article, a solution to placement problem of the various cells at the time of the realization of an integrated circuit is proposed by using the KOHONEN network.
Neural Networks ( ) Contents lists available at ScienceDirect
Bolton, McLean
Neural Networks ( ) Â Contents lists available at ScienceDirect Neural Networks journal homepage of an anatomically realistic neural network in rat vibrissal cortex Stefan Langa,b , Vincent J. Dercksenc , Bert electrical signals that propagate through anatomically realistic models of average neural networks
Reviews of computing technology: An overview of neural networks
Rainsford, A.E.
1992-02-15
This report discusses the historical background, models, computer hardware, and uses of neural networks. (LSP)
Neural Networks 1 Running Head: NETWORKS AND SPEED-ACCURACY TRADEOFF
Neural Networks 1 Running Head: NETWORKS AND SPEED-ACCURACY TRADEOFF Neural Networks Associated@columbia.edu (Y. Stern) #12;Neural Networks 2 Abstract This functional neuroimaging (fMRI) study examined the neural networks (spatial patterns of covarying neural activity) associated with the speed
Tampa Electric Neural Network Sootblowing
Mark A. Rhode
2004-09-30
Boiler combustion dynamics change continuously due to several factors including coal quality, boiler loading, ambient conditions, changes in slag/soot deposits and the condition of plant equipment. NOx formation, Particulate Matter (PM) emissions, and boiler thermal performance are directly affected by the sootblowing practices on a unit. As part of its Power Plant Improvement Initiative program, the US DOE is providing cofunding (DE-FC26-02NT41425) and NETL is the managing agency for this project at Tampa Electric's Big Bend Station. This program serves to co-fund projects that have the potential to increase thermal efficiency and reduce emissions from coal-fired utility boilers. A review of the Big Bend units helped identify intelligent sootblowing as a suitable application to achieve the desired objectives. The existing sootblower control philosophy uses sequential schemes, whose frequency is either dictated by the control room operator or is timed based. The intent of this project is to implement a neural network based intelligent sootblowing system, in conjunction with state-of-the-art controls and instrumentation, to optimize the operation of a utility boiler and systematically control boiler fouling. Utilizing unique, on-line, adaptive technology, operation of the sootblowers can be dynamically controlled based on real-time events and conditions within the boiler. This could be an extremely cost-effective technology, which has the ability to be readily and easily adapted to virtually any pulverized coal fired boiler. Through unique on-line adaptive technology, Neural Network-based systems optimize the boiler operation by accommodating equipment performance changes due to wear and maintenance activities, adjusting to fluctuations in fuel quality, and improving operating flexibility. The system dynamically adjusts combustion setpoints and bias settings in closed-loop supervisory control to simultaneously reduce NO{sub x} emissions and improve heat rate around the clock.
In Neural Networks, vol. 1, S1, p.552, 1988. ON THE EXPEDIENT USE OF NEURAL NETWORKS. Tony
Martinez, Tony R.
In Neural Networks, vol. 1, S1, p.552, 1988. ON THE EXPEDIENT USE OF NEURAL NETWORKS. Tony Martinez computer, calculators, special purpose logic devices, neural networks, etc. Each differs in its mechanism applications. Neural network features include parallel execution, adaptive learning, generalization, etc
Parametrizing Compton form factors with neural networks
Kresimir Kumericki; Dieter Mueller; Andreas Schafer
2011-12-08
We describe a method, based on neural networks, of revealing Compton form factors in the deeply virtual region. We compare this approach to standard least-squares model fitting both for a simplified toy case and for HERMES data.
Paraphrastic recurrent neural network language models
Liu, X.; Chen, X.; Gales, M. J. F.; Woodland, P. C.
2015-04-22
), “Hierarchical probabilistic neural network language model,” in Proc. International work- shop on artificial intelligence and statistics, Barbados, 2005, pp. 246–252. [24] M. Mohri (1997), “Finite-state transducers in language and speech processing...
Imbibition well stimulation via neural network design
Weiss, William (Socorro, NM)
2007-08-14
A method for stimulation of hydrocarbon production via imbibition by utilization of surfactants. The method includes use of fuzzy logic and neural network architecture constructs to determine surfactant use.
Computational Modeling of Neural Plasticity for Self-Organization of Neural Networks
Jin, Yaochu
Computational Modeling of Neural Plasticity for Self-Organization of Neural Networks Joseph Chrol on the learning per- formance of neural networks for accomplishing machine learning tasks such as classication, dynamics and learning per- formance of neural networks remains elusive. The purpose of this article
Modified high-order neural network for invariant pattern recognition
. Invariant pat- tern recognition using neural networks is a partic- ularly attractive approach because of its;bystructureisachievedbythestructureoftheneural network; good examples are the recognition (Fuku- shima, 2001) and high-order neural networks
Neural Networks and Fuzzy Systems for Nonlinear Applications
Wilamowski, Bogdan Maciej
Neural Networks and Fuzzy Systems for Nonlinear Applications Bogdan M. Wilamowski Director of with methods of computational intelligence such as neural networks and fuzzy systems. The problem such as neural networks and fuzzy systems. However, developments of neural or fuzzy systems are not trivial
Computation of Normal Logic Programs by Fibring Neural Networks
Seda, Anthony Karel
Computation of Normal Logic Programs by Fibring Neural Networks Vladimir Komendantsky1 and Anthony of the integration of fibring neural net- works (a generalization of conventional neural networks) into model by fibring neural networks of semantic immediate consequence operators TP and TP , where TP denotes
Deep Belief Networks The New Generation of Neural Networks1
Hernández Lobato, José Miguel
Deep Belief Networks The New Generation of Neural Networks1 Jos´e Miguel Hern´andez Lobato This presentation is mainly based on the work by Geoffrey E. Hinton. 1 / 28 #12;Deep Belief Networks Outline 1 Boltzmann Machines 2 Restricted Boltzmann Machines 3 Deep Belief Networks 4 Applications of Deep Belief
Williams, Chris
networks Christopher K. I. Williams Neural Computing Research Group Department of Computer Science Abstract For neural networks with a wide class of weight priors, it can be shown that in the limit a neural network by maximizing the likelihood of a finite amount of data it makes no sense to use a network
Generating Coherent Patterns of Activity from Chaotic Neural Networks
Abbott, Laurence
Neuron Article Generating Coherent Patterns of Activity from Chaotic Neural Networks David Sussillo be used to alter the chaotic activity of a recurrently connected neural network and generate complex but controlled outputs. Training a neural network is a process through which network parameters (typically
Neural Networks in Control Systems Theever-increasingtechnologicalde-
Antsaklis, Panos
Neural Networks in Control Systems Theever-increasingtechnologicalde- mands of ourmodem society require inno- vative approaches to highly demanding con- trol problems. Artificial neural networks, the controlcommunity has heardof neural networks and wonders if these networks can be used to provide better control
Natural Language Grammatical Inference with Recurrent Neural Networks
Giles, C. Lee
Natural Language Grammatical Inference with Recurrent Neural Networks Steve Lawrence, Member, IEEE of a complex grammar with neural networksÐspecifically, the task considered is that of training a network-and-Binding theory. Neural networks are trained, without the division into learned vs. innate components assumed
Designing asymmetric neural networks with associative memory
Zhao Hong [Department of Physics, Xiamen University, Xiamen 361005 (China)
2004-12-01
A strategy for designing asymmetric neural networks of associative memory with controllable degree of symmetry and controllable basins of attraction is presented. It is shown that the performance of the networks depends on the degree of the symmetry, and by adjusting the degree of the symmetry the spurious memories or unwanted attractors can be suppressed completely.
PSO assisted NURB neural network identification
Chen, Sheng
PSO assisted NURB neural network identification X. Hong1 and S. Chen2 1 School of Systems the shaping parameters in NURB network are esti- mated using a particle swarm optimization (PSO) procedure in the Hammerstein system. #12;II The PSO [15,16] constitutes a population based stochastic optimisation tech- nique
Modeling of a continuous food process with neural networks
Bullock, David Cole
1995-01-01
Three neural networks were constructed and trained to provide both next step prediction and multi-step prediction of a snack food continuous frying operation. The three neural models were a feedforward sigmoidal network ...
Pan Danguang; Gao Yanhua; Song Junlei [School of Civil and Environmental Engineering, University of Science and Technology Beijing, Beijing, 100083 (China)
2010-05-21
A new analysis technique, called multi-level interval estimation method, is developed for locating damage in structures. In this method, the artificial neural networks (ANN) analysis method is combined with the statistics theory to estimate the range of damage location. The ANN is multilayer perceptron trained by back-propagation. Natural frequencies and modal shape at a few selected points are used as input to identify the location and severity of damage. Considering the large-scale structures which have lots of elements, multi-level interval estimation method is developed to reduce the estimation range of damage location step-by-step. Every step, estimation range of damage location is obtained from the output of ANN by using the method of interval estimation. The next ANN training cases are selected from the estimation range after linear transform, and the output of new ANN estimation range of damage location will gained a reduced estimation range. Two numerical example analyses on 10-bar truss and 100-bar truss are presented to demonstrate the effectiveness of the proposed method.
Deep Neural Networks for Acoustic Modeling in Speech Recognition
Cortes, Corinna
1 Deep Neural Networks for Acoustic Modeling in Speech Recognition Geoffrey Hinton, Li Deng, Dong states as output. Deep neural networks with many hidden layers, that are trained using new methods have views of four research groups who have had recent successes in using deep neural networks for acoustic
Neural network model of creep strength of austenitic stainless steels
Cambridge, University of
Neural network model of creep strength of austenitic stainless steels T. Sourmail, H. K. D. H, and solution treatment temperature. The method involved a neural network analysis of a vast and general, and stress. Neural networks represent a more general regression method, which ameliorates most
Equivalence in Knowledge Representation: Automata, Recurrent Neural Networks,
Giles, C. Lee
Equivalence in Knowledge Representation: Automata, Recurrent Neural Networks, and Dynamical Fuzzy finite state automata (DFA) can be synthesized by or mapped into recurrent neural networks by directly programming the DFA structure into the weights of the neural network. Based on those results, a synthesis
Estimating uncertainty of streamflow simulation using Bayesian neural networks
Liang, Faming
Estimating uncertainty of streamflow simulation using Bayesian neural networks Xuesong Zhang,1 neural networks (BNNs) are powerful tools for providing reliable hydrologic prediction and quantifying of the uncertainties related to parameters (neural network's weights) and model structures were applied for uncertainty
Neural Networks: Tricks of the Trade R. C. Dimitriu
Cambridge, University of
Neural Networks: Tricks of the Trade R. C. Dimitriu 1 Data The first thing necessary to make a reliable neural network model is good quality data which are physically meaningful. It is also necessary in selecting the kind of data used in developing the neural network. An assessment should be made
A Neural Network for Locating the Primary Vertex in
Kantowski, Ron
A Neural Network for Locating the Primary Vertex in a Pixel Detector R. Kantowski and Caren Marzban, a neural network is trained to construct the coordinate of the primary vertex to a high degree of accuracy. Three other estimates of this coordinate are also considered and compared to that of the neural network
Neural networks in neuroscience: a brief overview Samuel Johnson1
Johnson, Samuel
1 Neural networks in neuroscience: a brief overview Samuel Johnson1 Instituto Carlos I de Física known as thought. However, the concept of a neural network (as understood in theoretical Hopfield in 1982 for his model of a neural network [2]. The situation considered was basically an Ising
Neural Networks www.biostat.wisc.edu/~dpage/cs760/
Page Jr., C. David
Neural Networks www.biostat.wisc.edu/~dpage/cs760/ 1 #12;Goals for the lecture you should units · multilayer neural networks · gradient descent · stochastic (online) gradient descent · sigmoid · early stopping · the role of hidden units · input encodings for neural networks · output encodings
Using Neural Networks to Forecast Stock Market Prices Ramon Lawrence
Lawrence, Ramon
Using Neural Networks to Forecast Stock Market Prices Ramon Lawrence Department of Computer Science on the application of neural networks in forecasting stock market prices. With their ability to discover patterns in nonlinear and chaotic systems, neural networks offer the ability to predict market directions more
Use of Neural Networks with Advection-Diffusion-Reaction Models
Hawai'i at Manoa, University of
Use of Neural Networks with Advection-Diffusion-Reaction Models to Estimate Large-Scale Movements-350 #12;Adam, M. S., and J. R. Sibert, Use of Neural Networks with Advection-Diffusion-Reaction Models Motivation 2 2 The model 2 3 Parameterizing movement fields 4 3.1 User of neural networks 5 3.2 Scaling
May 27, 2002 An Introduction to Neural Networks
Tesfatsion, Leigh
May 27, 2002 An Introduction to Neural Networks Vincent Cheung Kevin Cannons Signal & Data Advisor: Dr. W. Kinsner #12;Cheung/Cannons 1 Neural Networks Outline Fundamentals Classes Design and Verification Results and Discussion Conclusion #12;Cheung/Cannons 2 Neural Networks What Are Artificial
Neural Networks for Breast Cancer Diagnosis School of Computer Science
Yao, Xin
Neural Networks for Breast Cancer Diagnosis Xin Yao School of Computer Science The University neural network based approaches to breast cancer diag- nosis, both of which have displayed good general- isation. The first approach is based on evolution- ary artificial neural networks. In this approach
The Computational Power of Interactive Recur-rent Neural Networks
Siegelmann , Hava T
1 The Computational Power of Interactive Recur- rent Neural Networks J´er´emie Cabessa1 and Hava T, interactive computation, analog computation, re- current neural networks, interactive Turing machines-weighted recurrent neural networks were shown to be respectively equivalent to and strictly more powerful than
Estimating Uncertainty of Streamflow Simulation using Bayesian Neural Networks
Estimating Uncertainty of Streamflow Simulation using Bayesian Neural Networks Xuesong Zhang1-2607 Email: r-srinivasan@tamu.edu 1 #12;Abstract: Recent studies have shown that Bayesian Neural Networks of the uncertainties related to parameters (neural network's weights) and model structures were applied for uncertainty
Safety Criteria and Safety Lifecycle for Artificial Neural Networks
Kelly, Tim
Safety Criteria and Safety Lifecycle for Artificial Neural Networks Zeshan Kurd, Tim Kelly and Jim performance based techniques that aim to improve the safety of neural networks for safety critical for safety assurance. As a result, neural networks are typically restricted to advisory roles in safety
Neural Network forecasts of the tropical Pacific sea surface temperatures
Hsieh, William
Neural Network forecasts of the tropical Pacific sea surface temperatures Aiming Wu, William W Tang Jet Propulsion Laboratory, Pasadena, CA, USA Neural Networks (in press) December 11, 2005 title: Forecast of sea surface temperature 1 #12;Neural Network forecasts of the tropical Pacific sea
FAILURE BEHAVIOR IDENTIFICATION FOR A SPACE ANTENNA VIA NEURAL NETWORKS
Antsaklis, Panos
FAILURE BEHAVIOR IDENTIFICATION FOR A SPACE ANTENNA VIA NEURAL NETWORKS Michcl A. Sartoi David Taylo Research Center Code 1941 BeteA Mayln 200845000 ABSTRACT Using neural networks, a method-layer perceptron, the second uses a multi-layer perceptron and neural networks trained with the quadratic
Extracting Provably Correct Rules from Artificial Neural Networks
Clausen, Michael
Extracting Provably Correct Rules from Artificial Neural Networks Sebastian B. Thrun University procedures have been applied successfully to a variety of realworld scenarios, artificial neural networks that automatically compile neural networks into symbolic rules offer a promising perspective to overcome
Research and Implementation of Computer Simulation System for Neural Networks
Byrne, John H.
Research and Implementation of Computer Simulation System for Neural Networks Chen, Houjin Yuan: The ability of a neural network to process information and to generate a specific pattern of electrical computer simulation system for neural networks was designed and implemented including system architecture
Prediction of proteinprotein interaction sites in heterocomplexes with neural networks
Pazos, Florencio
Prediction of proteinprotein interaction sites in heterocomplexes with neural networks Piero on information about evolutionary con- servation and surface disposition. We implement a neural network based protein sur- face. However neural networks trained with a reduced representation of the interacting patch
Percolation Approach to Study Connectivity in Living Neural Networks
Moses, Elisha
Percolation Approach to Study Connectivity in Living Neural Networks Jordi Soriano, Ilan Breskin distribution and not a power law one. Keywords: neural networks, graphs, connectivity, percolation, giant as the fundamental feature to understand the potential of a living neural network. Unravelling the detailed
PPRODO: Prediction of Protein Domain Boundaries Using Neural Networks
Lee, Jooyoung
PPRODO: Prediction of Protein Domain Boundaries Using Neural Networks Jaehyun Sim, Seung-Yeon Kim-BLAST. A 10-fold cross-validation technique is performed to obtain the parameters of neural networks using; neural network INTRODUCTION Domains are semi-independent 3-dimensional (3D) units in proteins, and often
Prediction protein--protein interaction sites heterocomplexes with neural networks
Pazos, Florencio
Prediction protein--protein interaction sites heterocomplexes with neural networks Piero Fariselli neural network based system, which a cross validation proce dure and allows correct detection 73 face. However neural networks trained a reduced representation of interacting patch sequence profile su
A New Method for Mapping Optimization Problems onto Neural Networks
Peterson, Carsten
LU TP 891 March 1989 A New Method for Mapping Optimization Problems onto Neural Networks Carsten for obtaining approximate solutions to difficult optimization problems within the neural network paradigm redundancy that has plagued these problems when using straightforward neural network techniques
Safety Lifecycle for Developing Safety Critical Artificial Neural Networks
Kelly, Tim
Safety Lifecycle for Developing Safety Critical Artificial Neural Networks Zeshan Kurd, Tim Kelly.kelly}@cs.york.ac.uk Abstract. Artificial neural networks are employed in many areas of industry such as medicine and defence. There are many techniques that aim to improve the performance of neural networks for safety-critical systems
A Beginner's Guide to the Mathematics of Neural Networks
Coolen, ACC "Ton"
A Beginner's Guide to the Mathematics of Neural Networks A.C.C. Coolen Department of Mathematics ing our understanding of how neural networks operate, and the curious new mathematical conceptsexpert, I will present a biased selection of relatively simple examples of neural network tasks, models
Universal Distribution of Saliencies for Pruning in Layered Neural Networks
Lautrup, Benny
Universal Distribution of Saliencies for Pruning in Layered Neural Networks J. Gorodkin y, L. K learning in layered neural networks is a two stage process. A choice of architecture leads to the implicit is implementable in a neural network representation.12 xPermanent address: Department of Physiology and Institute
Forecasting Hot Water Consumption in Dwellings Using Artifitial Neural Networks
MacDonald, Mark
electricity consumption in time. This paper investigates the ability on Artificial Neural Networks to predict shift electric energy. Keywords--Hot Water Consumption; Forecasting; Artifitial Neural Networks; SmartForecasting Hot Water Consumption in Dwellings Using Artifitial Neural Networks Linas Gelazanskas
PROJECT REPORT USING NEURAL NETWORKS FOR APPROXIMATE RADIOSITY FORM FACTOR
Anderson, Charles W.
PROJECT REPORT USING NEURAL NETWORKS FOR APPROXIMATE RADIOSITY FORM FACTOR COMPUTATION Submitted, between each pair of objects within the scene. This project report explores the use of neural networks.3 Approximation with Neural Networks : : : : : : : : : : : : : : : : : : : : : : : 2 1.4 Outline of Project Report
EXPERT CONTROL BASED ON NEURAL NETWORKS FOR CONTROLLING
neural units in artificial neural networks (ANN), artificial neural network was used to develop a model of temperature model. The design of the network structure and transfer function was also discussed. Because growth in greenhouse climate. Temperature exerted significant effects on photosynthesis. The net
Chaotic time series prediction using artificial neural networks
Bartlett, E.B.
1991-12-31
This paper describes the use of artificial neural networks to model the complex oscillations defined by a chaotic Verhuist animal population dynamic. A predictive artificial neural network model is developed and tested, and results of computer simulations are given. These results show that the artificial neural network model predicts the chaotic time series with various initial conditions, growth parameters, or noise.
Chaotic time series prediction using artificial neural networks
Bartlett, E.B.
1991-01-01
This paper describes the use of artificial neural networks to model the complex oscillations defined by a chaotic Verhuist animal population dynamic. A predictive artificial neural network model is developed and tested, and results of computer simulations are given. These results show that the artificial neural network model predicts the chaotic time series with various initial conditions, growth parameters, or noise.
Dumidu Wijayasekara; Milos Manic; Piyush Sabharwall; Vivek Utgikar
2011-07-01
Artificial Neural Networks (ANN) have been used in the past to predict the performance of printed circuit heat exchangers (PCHE) with satisfactory accuracy. Typically published literature has focused on optimizing ANN using a training dataset to train the network and a testing dataset to evaluate it. Although this may produce outputs that agree with experimental results, there is a risk of over-training or overlearning the network rather than generalizing it, which should be the ultimate goal. An over-trained network is able to produce good results with the training dataset but fails when new datasets with subtle changes are introduced. In this paper we present EBaLM-OTR (error back propagation and Levenberg-Marquardt algorithms for over training resilience) technique, which is based on a previously discussed method of selecting neural network architecture that uses a separate validation set to evaluate different network architectures based on mean square error (MSE), and standard deviation of MSE. The method uses k-fold cross validation. Therefore in order to select the optimal architecture for the problem, the dataset is divided into three parts which are used to train, validate and test each network architecture. Then each architecture is evaluated according to their generalization capability and capability to conform to original data. The method proved to be a comprehensive tool in identifying the weaknesses and advantages of different network architectures. The method also highlighted the fact that the architecture with the lowest training error is not always the most generalized and therefore not the optimal. Using the method the testing error achieved was in the order of magnitude of within 10{sup -5} - 10{sup -3}. It was also show that the absolute error achieved by EBaLM-OTR was an order of magnitude better than the lowest error achieved by EBaLM-THP.
Cangelosi, Angelo
Cell division and migration in a 'genotype' for neural networks (Cell division and migration in neural networks) Angelo Cangelosi Domenico Parisi Stefano Nolfi Institute of Psychology - CNR 15, viale@gracco.irmkant.cnr.it email stefano@kant.irmkant.cnr.it In press in Network: computation in neural systems #12;1 Cell division
McLachlan, Geoff
738 IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 15, NO. 3, MAY 2004 Using the EM Algorithm to Train Neural Networks: Misconceptions and a New Algorithm for Multiclass Classification Shu-Kay Ng and Geoffrey in recent years as the basis for var- ious algorithms in application areas of neural networks such as pat
Toward IMRT 2D dose modeling using artificial neural networks: A feasibility study
Kalantzis, Georgios; Vasquez-Quino, Luis A.; Zalman, Travis; Pratx, Guillem; Lei, Yu [Radiation Oncology Department, University of Texas, Health Science Center San Antonio, Texas 78229 and Radiation Oncology Department, Stanford University School of Medicine, Stanford, California 94305 (United States); Radiation Oncology Department, University of Texas, Health Science Center San Antonio, Texas 78229 (United States); Radiation Oncology Department, Stanford University School of Medicine, Stanford, California 94305 (United States); Radiation Oncology Department, University of Texas, Health Science Center San Antonio, Texas 78229 (United States)
2011-10-15
Purpose: To investigate the feasibility of artificial neural networks (ANN) to reconstruct dose maps for intensity modulated radiation treatment (IMRT) fields compared with those of the treatment planning system (TPS). Methods: An artificial feed forward neural network and the back-propagation learning algorithm have been used to replicate dose calculations of IMRT fields obtained from PINNACLE{sup 3} v9.0. The ANN was trained with fluence and dose maps of IMRT fields for 6 MV x-rays, which were obtained from the amorphous silicon (a-Si) electronic portal imaging device of Novalis TX. Those fluence distributions were imported to the TPS and the dose maps were calculated on the horizontal midpoint plane of a water equivalent homogeneous cylindrical virtual phantom. Each exported 2D dose distribution from the TPS was classified into two clusters of high and low dose regions, respectively, based on the K-means algorithm and the Euclidian metric in the fluence-dose domain. The data of each cluster were divided into two sets for the training and validation phase of the ANN, respectively. After the completion of the ANN training phase, 2D dose maps were reconstructed by the ANN and isodose distributions were created. The dose maps reconstructed by ANN were evaluated and compared with the TPS, where the mean absolute deviation of the dose and the {gamma}-index were used. Results: A good agreement between the doses calculated from the TPS and the trained ANN was achieved. In particular, an average relative dosimetric difference of 4.6% and an average {gamma}-index passing rate of 93% were obtained for low dose regions, and a dosimetric difference of 2.3% and an average {gamma}-index passing rate of 97% for high dose region. Conclusions: An artificial neural network has been developed to convert fluence maps to corresponding dose maps. The feasibility and potential of an artificial neural network to replicate complex convolution kernels in the TPS for IMRT dose calculations have been demonstrated.
Auto-associative nanoelectronic neural network
Nogueira, C. P. S. M.; Guimarães, J. G. [Departamento de Engenharia Elétrica - Laboratório de Dispositivos e Circuito Integrado, Universidade de Brasília, CP 4386, CEP 70904-970 Brasília DF (Brazil)
2014-05-15
In this paper, an auto-associative neural network using single-electron tunneling (SET) devices is proposed and simulated at low temperature. The nanoelectronic auto-associative network is able to converge to a stable state, previously stored during training. The recognition of the pattern involves decreasing the energy of the input state until it achieves a point of local minimum energy, which corresponds to one of the stored patterns.
Bayesian and Maximum Likelihood Neural Networks
Nielsen, Finn Årup
Bayesian and Maximum Likelihood Neural Networks Finn A ffi rup Nielsen Section for Digital Signal, linear output, Gaussian distribution ] \\Gamma 1;+1[ ffl Binary (binary classification), tanh on output, bino mial distribution. ] \\Gamma 1; +1[ ffl Classification, softmax function on outputs [Bridle, 1990
Paraphrastic Neural Network Language Models
Liu, X.; Gales, M. J. F.; Woodland, P. C.
2014-01-01
Intelligence and Statistics, Barbados, 2005, pp.246-252. [24] M. Mohri (1997). “Finite-state transducers in language and speech processing”, Computational Linguistics, 23:2, 1997. [25] J. Park, X. Liu, M. J. F. Gales and P. C. Woodland (2010), “Improved neural...
Adaptive Neural Networks for Automatic Negotiation
Sakas, D. P.; Vlachos, D. S.; Simos, T. E. [University of Peloponnese, 22100 Tripoli (Greece)
2007-12-26
The use of fuzzy logic and fuzzy neural networks has been found effective for the modelling of the uncertain relations between the parameters of a negotiation procedure. The problem with these configurations is that they are static, that is, any new knowledge from theory or experiment lead to the construction of entirely new models. To overcome this difficulty, we apply in this work, an adaptive neural topology to model the negotiation process. Finally a simple simulation is carried in order to test the new method.
Informatica 17 page xxx--yyy 1 ON BAYESIAN NEURAL NETWORKS
Kononenko, Igor
Informatica 17 page xxx--yyy 1 ON BAYESIAN NEURAL NETWORKS Igor Kononenko University of Ljubljana neural network, Hopfield's neural network, naive Bayesian classifier, con tinuous neural network: ??? Revised: ??? Accepted: ??? In the paper the contribution of the work on Bayesian neural networks
Polarized DIS Structure Functions from Neural Networks
Del Debbio, L.; Guffanti, A. [School of Physics, University of Edinburgh, Edinburgh (United Kingdom); Piccione, A. [Universita degli Studi di Torino, Torino (Italy); INFN Torino, Torino (Italy)
2007-06-13
We present a parametrization of polarized Deep-Inelastic-Scattering (DIS) structure functions based on Neural Networks. The parametrization provides a bias-free determination of the probability measure in the space of structure functions, which retains information on experimental errors and correlations. As an example we discuss the application of this method to the study of the structure function g{sub 1}{sup p}(x,Q{sup 2})
Conceptualization and image understanding by neural networks
Gudipalley, Chandu
1993-01-01
. Traditional roles of artificial neural networks have been restricted to pattern recognition (feature extraction, segmentation and classification) as opposed to pattern understanding (interpretation of the identified patterns to per- form a specific task... Conceptualization and Fuzzy Semantics 2. Conceptual Associations 3. Pattern-Concept Association 4. Concept-Pattern Association 5. Concept-Concept Association 6. Conceptual Analysis 7. Conceptualization Depth vs Resolution Depth . 8. Focusing of Attention 9...
Artificial neural network cardiopulmonary modeling and diagnosis
Kangas, Lars J. (Richland, WA); Keller, Paul E. (Richland, WA)
1997-01-01
The present invention is a method of diagnosing a cardiopulmonary condition in an individual by comparing data from a progressive multi-stage test for the individual to a non-linear multi-variate model, preferably a recurrent artificial neural network having sensor fusion. The present invention relies on a cardiovascular model developed from physiological measurements of an individual. Any differences between the modeled parameters and the parameters of an individual at a given time are used for diagnosis.
Artificial neural network cardiopulmonary modeling and diagnosis
Kangas, L.J.; Keller, P.E.
1997-10-28
The present invention is a method of diagnosing a cardiopulmonary condition in an individual by comparing data from a progressive multi-stage test for the individual to a non-linear multi-variate model, preferably a recurrent artificial neural network having sensor fusion. The present invention relies on a cardiovascular model developed from physiological measurements of an individual. Any differences between the modeled parameters and the parameters of an individual at a given time are used for diagnosis. 12 figs.
Analysis of complex systems using neural networks
Uhrig, R.E. [Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering]|[Oak Ridge National Lab., TN (United States)
1992-12-31
The application of neural networks, alone or in conjunction with other advanced technologies (expert systems, fuzzy logic, and/or genetic algorithms), to some of the problems of complex engineering systems has the potential to enhance the safety, reliability, and operability of these systems. Typically, the measured variables from the systems are analog variables that must be sampled and normalized to expected peak values before they are introduced into neural networks. Often data must be processed to put it into a form more acceptable to the neural network (e.g., a fast Fourier transformation of the time-series data to produce a spectral plot of the data). Specific applications described include: (1) Diagnostics: State of the Plant (2) Hybrid System for Transient Identification, (3) Sensor Validation, (4) Plant-Wide Monitoring, (5) Monitoring of Performance and Efficiency, and (6) Analysis of Vibrations. Although specific examples described deal with nuclear power plants or their subsystems, the techniques described can be applied to a wide variety of complex engineering systems.
Analysis of complex systems using neural networks
Uhrig, R.E. (Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering Oak Ridge National Lab., TN (United States))
1992-01-01
The application of neural networks, alone or in conjunction with other advanced technologies (expert systems, fuzzy logic, and/or genetic algorithms), to some of the problems of complex engineering systems has the potential to enhance the safety, reliability, and operability of these systems. Typically, the measured variables from the systems are analog variables that must be sampled and normalized to expected peak values before they are introduced into neural networks. Often data must be processed to put it into a form more acceptable to the neural network (e.g., a fast Fourier transformation of the time-series data to produce a spectral plot of the data). Specific applications described include: (1) Diagnostics: State of the Plant (2) Hybrid System for Transient Identification, (3) Sensor Validation, (4) Plant-Wide Monitoring, (5) Monitoring of Performance and Efficiency, and (6) Analysis of Vibrations. Although specific examples described deal with nuclear power plants or their subsystems, the techniques described can be applied to a wide variety of complex engineering systems.
Dynamical synapses causing self-organized criticality in neural networks
Loss, Daniel
LETTERS Dynamical synapses causing self-organized criticality in neural networks A. LEVINA1,2,3 , J more realistic) dynamical synapses14 in a spiking neural network, the neuronal avalanches turn from dynamics is robust to parameter changes. Consider a network of N integrate-and-fire neurons. Each neuron
Evolution of Neural Networks for Helicopter Control: Why Modularity Matters
Togelius, Julian
Evolution of Neural Networks for Helicopter Control: Why Modularity Matters Renzo De Nardi, Julian helicopter flocking. A methodology is proposed in which neural network based controllers are evolved in a simulation using a dynamic model qualitatively similar to the physical helicopter. Several network
Neural node network and model, and method of teaching same
Parlos, Alexander G. (College Station, TX); Atiya, Amir F. (College Station, TX); Fernandez, Benito (Austin, TX); Tsai, Wei K. (Irvine, CA); Chong, Kil T. (College Station, TX)
1995-01-01
The present invention is a fully connected feed forward network that includes at least one hidden layer 16. The hidden layer 16 includes nodes 20 in which the output of the node is fed back to that node as an input with a unit delay produced by a delay device 24 occurring in the feedback path 22 (local feedback). Each node within each layer also receives a delayed output (crosstalk) produced by a delay unit 36 from all the other nodes within the same layer 16. The node performs a transfer function operation based on the inputs from the previous layer and the delayed outputs. The network can be implemented as analog or digital or within a general purpose processor. Two teaching methods can be used: (1) back propagation of weight calculation that includes the local feedback and the crosstalk or (2) more preferably a feed forward gradient decent which immediately follows the output computations and which also includes the local feedback and the crosstalk. Subsequent to the gradient propagation, the weights can be normalized, thereby preventing convergence to a local optimum. Education of the network can be incremental both on and off-line. An educated network is suitable for modeling and controlling dynamic nonlinear systems and time series systems and predicting the outputs as well as hidden states and parameters. The educated network can also be further educated during on-line processing.
Neural node network and model, and method of teaching same
Parlos, A.G.; Atiya, A.F.; Fernandez, B.; Tsai, W.K.; Chong, K.T.
1995-12-26
The present invention is a fully connected feed forward network that includes at least one hidden layer. The hidden layer includes nodes in which the output of the node is fed back to that node as an input with a unit delay produced by a delay device occurring in the feedback path (local feedback). Each node within each layer also receives a delayed output (crosstalk) produced by a delay unit from all the other nodes within the same layer. The node performs a transfer function operation based on the inputs from the previous layer and the delayed outputs. The network can be implemented as analog or digital or within a general purpose processor. Two teaching methods can be used: (1) back propagation of weight calculation that includes the local feedback and the crosstalk or (2) more preferably a feed forward gradient decent which immediately follows the output computations and which also includes the local feedback and the crosstalk. Subsequent to the gradient propagation, the weights can be normalized, thereby preventing convergence to a local optimum. Education of the network can be incremental both on and off-line. An educated network is suitable for modeling and controlling dynamic nonlinear systems and time series systems and predicting the outputs as well as hidden states and parameters. The educated network can also be further educated during on-line processing. 21 figs.
Neural networks and their application to nuclear power plant diagnosis
Reifman, J. [Argonne National Lab., IL (United States). Reactor Analysis Div.
1997-10-01
The authors present a survey of artificial neural network-based computer systems that have been proposed over the last decade for the detection and identification of component faults in thermal-hydraulic systems of nuclear power plants. The capabilities and advantages of applying neural networks as decision support systems for nuclear power plant operators and their inherent characteristics are discussed along with their limitations and drawbacks. The types of neural network structures used and their applications are described and the issues of process diagnosis and neural network-based diagnostic systems are identified. A total of thirty-four publications are reviewed.
R. Rojas: Neural Networks, Springer-Verlag, Berlin, 1996 Raul Rojas
Block, Marco
R. Rojas: Neural Networks, Springer-Verlag, Berlin, 1996 Ra´ul Rojas Neural Networks A Systematic Introduction Springer Berlin Heidelberg NewYork Hong Kong London Milan Paris Tokyo R. Rojas: Neural Networks, Springer-Verlag, Berlin, 1996R. Rojas: Neural Networks, Springer-Verlag, Berlin, 1996R. Rojas: Neural
Neural Networks 23 (2010) 667668 Contents lists available at ScienceDirect
Fukai, Tomoki
2010-01-01
Neural Networks 23 (2010) 667Â668 Contents lists available at ScienceDirect Neural Networks journal of large-scale neural network simulations poses the same challenge as experimental data does. This special massively parallel neural recordings and large-scale neural network simulations; (b) Frameworks
Desynchronization in diluted neural networks
Zillmer, Ruediger [INFN Sezione Firenze, via Sansone 1, I-50019 Sesto Fiorentino (Italy); Livi, Roberto [Dipartimento di Fisica, Universita di Firenze, via Sansone 1, I-50019 Sesto Fiorentino (Italy); Sezione INFN, Unita' INFM e Centro Interdipartimentale per lo Studio delle Dinamiche Complesse, via Sansone 1, I-50019 Sesto Fiorentino (Italy); Politi, Antonio; Torcini, Alessandro [Istituto dei Sistemi Complessi, CNR, CNR, via Madonna del Piano 10, I-50019 Sesto Fiorentino (Italy); Centro Interdipartimentale per lo Studio delle Dinamiche Complesse, via Sansone 1, I-50019 Sesto Fiorentino (Italy)
2006-09-15
The dynamical behavior of a weakly diluted fully inhibitory network of pulse-coupled spiking neurons is investigated. Upon increasing the coupling strength, a transition from regular to stochasticlike regime is observed. In the weak-coupling phase, a periodic dynamics is rapidly approached, with all neurons firing with the same rate and mutually phase locked. The strong-coupling phase is characterized by an irregular pattern, even though the maximum Lyapunov exponent is negative. The paradox is solved by drawing an analogy with the phenomenon of 'stable chaos', i.e., by observing that the stochasticlike behavior is 'limited' to an exponentially long (with the system size) transient. Remarkably, the transient dynamics turns out to be stationary.
Artificial Neural Network Circuit for Spectral Pattern Recognition
Rasheed, Farah
2013-09-04
Artificial Neural Networks (ANNs) are a massively parallel network of a large number of interconnected neurons similar to the structure of biological neurons in the human brain. ANNs find applications in a large number of fields, from pattern...
Applications of Neural Networks in Hadron Physics
Krzysztof M. Graczyk; Cezary Juszczak
2014-09-18
The Bayesian approach for the feed-forward neural networks is reviewed. Its potential for usage in hadron physics is discussed. As an example of the application the study of the the two-photon exchange effect is presented. We focus on the model comparison, the estimation of the systematic uncertainties due to the choice of the model, and the over-fitting. As an illustration the predictions of the cross sections ratio $d \\sigma(e^+ p\\to e^+ p)/d \\sigma(e^- p\\to e^- p)$ are given together with the estimate of the uncertainty due to the parametrization choice.
NEURObjects: an object-oriented library for neural network development
Masulli, Francesco
-Oriented (OO) program- ming (see, e.g., the libraries of the Timothy Masters's book [22]). Others needNEURObjects: an object-oriented library for neural network development Giorgio Valentini@disi.unige.it, masulli@disi.unige.it Abstract NEURObjects is a set of C++ library classes for neural network development
Developmental Plasticity in Cartesian Genetic Programming Artificial Neural Networks
Fernandez, Thomas
Developmental Plasticity in Cartesian Genetic Programming Artificial Neural Networks Maryam Mahsal developmental plasticity in Artificial Neural Networks using Carte- sian Genetic Programming. This is inspired by developmental plasticity that exists in the biological brain allowing it to adapt to a changing environment
A New Stochastic PSO Technique for Neural Network Training
Li, Yangmin
A New Stochastic PSO Technique for Neural Network Training Yangmin Li and Xin Chen Department Optimization(PSO) has been widely applied for training neural network. To improve the performance of PSO paradigm of particle swarm optimization named stochastic PSO (S-PSO). The feature of the S-PSO is its high
Ungar, Lyle H.
Selected Neural Networks Bibliography As there are dozens of books and tens of thousands of articles on neural networks, this does not try to be comprehensive. It should, however, provide a good@cis.upenn.edu (215) 8987449 Introduction to Neural Networks Books Ripley, B.D., ``Pattern Recognition and Neural
Spikes Synchronization in Neural Networks with Synaptic Plasticity
Borges, Rafael R; Batista, Antonio M; Caldas, Iberê L; Borges, Fernando S; Lameu, Ewandson L
2015-01-01
In this paper, we investigated the neural spikes synchronisation in a neural network with synaptic plasticity and external perturbation. In the simulations the neural dynamics is described by the Hodgkin Huxley model considering chemical synapses (excitatory) among neurons. According to neural spikes synchronisation is expected that a perturbation produce non synchronised regimes. However, in the literature there are works showing that the combination of synaptic plasticity and external perturbation may generate synchronised regime. This article describes the effect of the synaptic plasticity on the synchronisation, where we consider a perturbation with a uniform distribution. This study is relevant to researches of neural disorders control.
6.2.8 Neural networks for data mining Walter Kosters
Kosters, Walter
6.2.8 Neural networks for data mining Walter Kosters 1 In many application areas neural networks distinctive member of the large family of neural networks), and we final- ly focus on their usefulness notice that the idea of neural networks originates from the physiology of the human brain. General
Financial Market Modeling with Quantum Neural Networks
Gonçalves, Carlos Pedro
2015-01-01
Econophysics has developed as a research field that applies the formalism of Statistical Mechanics and Quantum Mechanics to address Economics and Finance problems. The branch of Econophysics that applies of Quantum Theory to Economics and Finance is called Quantum Econophysics. In Finance, Quantum Econophysics' contributions have ranged from option pricing to market dynamics modeling, behavioral finance and applications of Game Theory, integrating the empirical finding, from human decision analysis, that shows that nonlinear update rules in probabilities, leading to non-additive decision weights, can be computationally approached from quantum computation, with resulting quantum interference terms explaining the non-additive probabilities. The current work draws on these results to introduce new tools from Quantum Artificial Intelligence, namely Quantum Artificial Neural Networks as a way to build and simulate financial market models with adaptive selection of trading rules, leading to turbulence and excess ku...
Delayed switching applied to memristor neural networks
Wang, Frank Z.; Yang Xiao; Lim Guan [Future Computing Group, School of Computing, University of Kent, Canterbury (United Kingdom); Helian Na [School of Computer Science, University of Hertfordshire, Hatfield (United Kingdom); Wu Sining [Xyratex, Havant (United Kingdom); Guo Yike [Department of Computing, Imperial College, London (United Kingdom); Rashid, Md Mamunur [CERN, Geneva (Switzerland)
2012-04-01
Magnetic flux and electric charge are linked in a memristor. We reported recently that a memristor has a peculiar effect in which the switching takes place with a time delay because a memristor possesses a certain inertia. This effect was named the ''delayed switching effect.'' In this work, we elaborate on the importance of delayed switching in a brain-like computer using memristor neural networks. The effect is used to control the switching of a memristor synapse between two neurons that fire together (the Hebbian rule). A theoretical formula is found, and the design is verified by a simulation. We have also built an experimental setup consisting of electronic memristive synapses and electronic neurons.
Beneficial role of noise in artificial neural networks
Monterola, Christopher; Saloma, Caesar; Zapotocky, Martin
2008-06-18
We demonstrate enhancement of neural networks efficacy to recognize frequency encoded signals and/or to categorize spatial patterns of neural activity as a result of noise addition. For temporal information recovery, noise directly added to the receiving neurons allow instantaneous improvement of signal-to-noise ratio [Monterola and Saloma, Phys. Rev. Lett. 2002]. For spatial patterns however, recurrence is necessary to extend and homogenize the operating range of a feed-forward neural network [Monterola and Zapotocky, Phys. Rev. E 2005]. Finally, using the size of the basin of attraction of the networks learned patterns (dynamical fixed points), a procedure for estimating the optimal noise is demonstrated.
An introduction to genetic algorithms for neural networks
Cambridge, University of
An introduction to genetic algorithms for neural networks Richard Kemp 1 Introduction Once a neural can use a genetic algorithm to try and solve the problem. What are genetic algorithms? Genetic algorithms (GAs) are search algo- rithms based on the mechanics of natural selection and genetics as observed
Neural Network Approach to Locating Cryptography in Object Code
Jason L. Wright; Milos Manic
2009-09-01
Finding and identifying cryptography is a growing concern in the malware analysis community. In this paper, artificial neural networks are used to classify functional blocks from a disassembled program as being either cryptography related or not. The resulting system, referred to as NNLC (Neural Net for Locating Cryptography) is presented and results of applying this system to various libraries are described.
Learning, Memory, and the Role of Neural Network Architecture
Carlson, Jean
Learning, Memory, and the Role of Neural Network Architecture Ann M. Hermundstad*, Kevin S. Brown architecture. In this study, we compare the performance of parallel and layered network architectures during to complexity in network architecture by characterizing local error landscape curvature. We find that variations
EXPERIMENTAL ANALYSIS OF INPUT WEIGHT FREEZING IN CONSTRUCTIVE NEURAL NETWORKS
Yeung, Dit-Yan
EXPERIMENTAL ANALYSIS OF INPUT WEIGHT FREEZING IN CONSTRUCTIVE NEURAL NETWORKS TinÂYau Kwok Dit network. One calls for freezing the input weights of the original network and training only those of freezing is different for different problem domains and hence is not conclusive. This paper describes
RESERVOIR INFLOW FORECASTING USING NEURAL NETWORKS CHANDRASHEKAR SUBRAMANIAN
Manry, Michael
a mixture of hydroelectric and non- hydroelectric power, the economics of the hydroelectric plants depend, and to economically allocate the load between various non-hydroelectric plants. Neural networks provide an attractive
A neural network approach to snack quality evaluation
Sayeed, Mohammad Shaheen
1994-01-01
A neural network approach is investigated in this study for the evaluation of the quality of typical snack products. Although quality of food appears mostly subjective, some external features of the product are indicators of snack quality. External...
Neural Network Parameterizations of Electromagnetic Nucleon Form Factors
Krzysztof M. Graczyk; Piotr Plonski; Robert Sulej
2010-08-25
The electromagnetic nucleon form-factors data are studied with artificial feed forward neural networks. As a result the unbiased model-independent form-factor parametrizations are evaluated together with uncertainties. The Bayesian approach for the neural networks is adapted for chi2 error-like function and applied to the data analysis. The sequence of the feed forward neural networks with one hidden layer of units is considered. The given neural network represents a particular form-factor parametrization. The so-called evidence (the measure of how much the data favor given statistical model) is computed with the Bayesian framework and it is used to determine the best form factor parametrization.
Detection of Phonological Features in Continuous Speech using Neural Networks
King, Simon; Taylor, Paul
which uses binary features, 2) a multi valued (MV) feature system which uses traditional phonetic categories such as manner, place etc, and 3) Government Phonology (GP) which uses a set of structured primes. All experiments used recurrent neural networks...
Classification of fuels using multilayer perceptron neural networks
Ozaki, Sergio T. R.; Wiziack, Nadja K. L.; Paterno, Leonardo G.; Fonseca, Fernando J. [Department of Electronic Systems Engineering, Polytechnic School, University of Sao Paulo Avenida Professor Luciano Gualberto, travessa 3, 158, 05508-900, Sao Paulo-SP (Brazil)
2009-05-23
Electrical impedance data obtained with an array of conducting polymer chemical sensors was used by a neural network (ANN) to classify fuel adulteration. Real samples were classified with accuracy greater than 90% in two groups: approved and adulterated.
Monte Carlo event reconstruction implemented with artificial neural networks
Tolley, Emma Elizabeth
2011-01-01
I implemented event reconstruction of a Monte Carlo simulation using neural networks. The OLYMPUS Collaboration is using a Monte Carlo simulation of the OLYMPUS particle detector to evaluate systematics and reconstruct ...
Characterization of Shape Memory Alloys Using Artificial Neural Networks
Henrickson, James V
2014-04-28
shape memory alloy material parameters with satisfactory accuracy. Comparison of the implemented training data generation methods indicates that the Taguchi-based approach yields an artificial neural network that outperforms that of the factorial...
Model building in neural networks with hidden Markov models
Wynne-Jones, Michael
This thesis concerns the automatic generation of architectures for neural networks and other pattern recognition models comprising many elements of the same type. The requirement for such models, with automatically ...
Error bars for linear and nonlinear neural network regression models
Penny, Will
Error bars for linear and nonlinear neural network regression models William D. Penny and Stephen J College of Science, Technology and Medicine, London SW7 2BT., U.K. w.penny@ic.ac.uk, s
The Neural Network Method of Corrosion Diagnosis for Grounding Grid
Hou Zaien [School of Science, Shaanxi University of Sci. and Tech., Xi'an, 710021 (China); Duan Fujian [School of Science, Guilin University of Electronic Tech., Guilin, 541004 (China); Zhang Kecun [School of Science, Xi'an Jiaotong University, Xi'an, 710049 (China)
2008-11-06
Safety of persons, protection of equipment and continuity of power supply are the main objectives of the grounding system of a large electrical installation. For its accurate working status, it is essential to determine every branch resistance in the system. In this paper, we present a neural network method of corrosion diagnosis for the grounding grid based on the neural network theory. The feasibility of this method is discussed by means of its application to a simulant grounding grid.
AUTOMATED DEFECT CLASSIFICATION USING AN ARTIFICIAL NEURAL NETWORK
Chady, T.; Caryk, M. [Szczecin University of Technology, Department of Electrical Engineering (Poland); Piekarczyk, B. [Technic-Control, Szczecin (Poland)
2009-03-03
The automated defect classification algorithm based on artificial neural network with multilayer backpropagation structure was utilized. The selected features of flaws were used as input data. In order to train the neural network it is necessary to prepare learning data which is representative database of defects. Database preparation requires the following steps: image acquisition and pre-processing, image enhancement, defect detection and feature extraction. The real digital radiographs of welded parts of a ship were used for this purpose.
A portable neural network approach to vehicle tracking
Miller, Kelly Maxwell
1994-01-01
A PORTABLE NEURAL NETWORK APPROACH TO VEHICLE TRACKING A Thesis by KELLY MAXWELL MILLER Submitted to the Office of Graduate Studies of Texas A&M University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE May... 1994 Major Subject: Electrical Engineering A PORTABLE NEURAL NETWORK APPROACH TO VEHICLE TRACKING A Thesis by KELLY MAXWELL MILLER Submitted to the OAice of Graduate Studies of Texas ARM University in partial fulfillment of the requirements...
Hybrid digital signal processing and neural networks applications in PWRs
Eryurek, E.; Upadhyaya, B.R.; Kavaklioglu, K.
1991-01-01
Signal validation and plant subsystem tracking in power and process industries require the prediction of one or more state variables. Both heteroassociative and auotassociative neural networks were applied for characterizing relationships among sets of signals. A multi-layer neural network paradigm was applied for sensor and process monitoring in a Pressurized Water Reactor (PWR). This nonlinear interpolation technique was found to be very effective for these applications.
Hybrid digital signal processing and neural networks applications in PWRs
Eryurek, E.; Upadhyaya, B.R.; Kavaklioglu, K.
1991-12-31
Signal validation and plant subsystem tracking in power and process industries require the prediction of one or more state variables. Both heteroassociative and auotassociative neural networks were applied for characterizing relationships among sets of signals. A multi-layer neural network paradigm was applied for sensor and process monitoring in a Pressurized Water Reactor (PWR). This nonlinear interpolation technique was found to be very effective for these applications.
Coupling neural networks to incomplete dynamical systems via variational data assimilation
Hsieh, William
Coupling neural networks to incomplete dynamical systems via variational data assimilationforward neural network (NN) model opens the possibility of hybrid neuraldynamical models via variational data value decomposition (SVD) (Syu and Neelin 1995), or by a neural network (Tang et al. 1999). While
Neural Networks 24 (2011) 950960 Contents lists available at SciVerse ScienceDirect
Fukai, Tomoki
2011-01-01
Neural Networks 24 (2011) 950Â960 Contents lists available at SciVerse ScienceDirect Neural of a spiking neural network model of the basal ganglia circuitry using general purpose computing on graphics neural networks Real-time simulation GPGPUs Basal ganglia High-performance computing a b s t r a c t Real
Neural Network Based Intelligent Sootblowing System
Mark Rhode
2005-04-01
Cost effective generation of electricity is vital to the economic growth and stability of this nation. To accomplish this goal a balanced portfolio of fuel sources must be maintained and established which not only addresses the cost of conversion of these energy sources to electricity, but also does so in an efficient and environmentally sound manner. Conversion of coal as an energy source to produce steam for a variety of systems has been a cornerstone of modern industry. However, the use of coal in combustion systems has traditionally produced unacceptable levels of gaseous and particulate emissions, albeit that recent combustion, removal and mitigation techniques have drastically reduced these levels. With the combustion of coal there is always the formation and deposition of ash and slag within the boilers. This adversely affects the rate at which heat is transferred to the working fluid, which in the case of electric generators is water/steam. The fouling of the boiler leads to poor efficiencies due to the fact that heat which could normally be transferred to the working fluid remains in the flue gas stream and exits to the environment without beneficial use. This loss in efficiency translates to higher consumption of fuel for equivalent levels of electric generation; hence more gaseous emissions are also produced. Another less obvious problem exists with fouling of various sections of the boiler creating intense peak temperatures within and around the combustion zone. Total nitrogen oxides (NOx) generation is primarily a function of both ''fuel'' and ''thermal'' NOx production. Fuel NOx which generally comprises 20%-40% of the total NOx generated is predominantly influenced by the levels of oxygen present, while thermal NOx which comprises the balance is a function of temperature. As the fouling of the boiler increases and the rate of heat transfer decreases, peak temperatures increase as does the thermal NOx production. Due to the composition of coal, particulate matter is also a by-product of coal combustion. Modern day utility boilers are usually fitted with electrostatic precipitators to aid in the collection of particulate matter. Although extremely efficient, these devices are sensitive to rapid changes in inlet mass concentration as well as total mass loading. Traditionally, utility boilers are equipped with devices known as sootblowers, which use, steam, water or air to dislodge and clean the surfaces within the boiler and are operated based upon established rule or operator's judgment. Poor sootblowing regimes can influence particulate mass loading to the electrostatic precipitators. The project applied a neural network intelligent sootblowing system in conjunction with state-of-the-art controls and instruments to optimize the operation of a utility boiler and systematically control boiler slagging/fouling. This optimization process targeted reduction of NOx of 30%, improved efficiency of 2% and a reduction in opacity of 5%. The neural network system proved to be a non-invasive system which can readily be adapted to virtually any utility boiler. Specific conclusions from this neural network application are listed below. These conclusions should be used in conjunction with the specific details provided in the technical discussions of this report to develop a thorough understanding of the process.
Neural and Fuzzy Adaptive Control of Induction Motor Drives
Bensalem, Y. [Research Unit of Modelisation, Analyse, Command of Systems MACS (Tunisia); Sbita, L.; Abdelkrim, M. N. [6029 Universite High School of Engineering-Gabes-Tunisia (Tunisia)
2008-06-12
This paper proposes an adaptive neural network speed control scheme for an induction motor (IM) drive. The proposed scheme consists of an adaptive neural network identifier (ANNI) and an adaptive neural network controller (ANNC). For learning the quoted neural networks, a back propagation algorithm was used to automatically adjust the weights of the ANNI and ANNC in order to minimize the performance functions. Here, the ANNI can quickly estimate the plant parameters and the ANNC is used to provide on-line identification of the command and to produce a control force, such that the motor speed can accurately track the reference command. By combining artificial neural network techniques with fuzzy logic concept, a neural and fuzzy adaptive control scheme is developed. Fuzzy logic was used for the adaptation of the neural controller to improve the robustness of the generated command. The developed method is robust to load torque disturbance and the speed target variations when it ensures precise trajectory tracking with the prescribed dynamics. The algorithm was verified by simulation and the results obtained demonstrate the effectiveness of the IM designed controller.
Fast cosmological parameter estimation using neural networks
T. Auld; M. Bridges; M. P. Hobson; S. F. Gull
2007-09-17
We present a method for accelerating the calculation of CMB power spectra, matter power spectra and likelihood functions for use in cosmological parameter estimation. The algorithm, called CosmoNet, is based on training a multilayer perceptron neural network and shares all the advantages of the recently released Pico algorithm of Fendt & Wandelt, but has several additional benefits in terms of simplicity, computational speed, memory requirements and ease of training. We demonstrate the capabilities of CosmoNet by computing CMB power spectra over a box in the parameter space of flat \\Lambda CDM models containing the 3\\sigma WMAP1 confidence region. We also use CosmoNet to compute the WMAP3 likelihood for flat \\Lambda CDM models and show that marginalised posteriors on parameters derived are very similar to those obtained using CAMB and the WMAP3 code. We find that the average error in the power spectra is typically 2-3% of cosmic variance, and that CosmoNet is \\sim 7 \\times 10^4 faster than CAMB (for flat models) and \\sim 6 \\times 10^6 times faster than the official WMAP3 likelihood code. CosmoNet and an interface to CosmoMC are publically available at www.mrao.cam.ac.uk/software/cosmonet.
SE2NN11 Neural Networks Part B Dr Richard Mitchell, 2014 1
Mitchell, Richard
SE2NN11 Neural Networks Part B © Dr Richard Mitchell, 2014 1 p1 RJM 16/09/14 SE2NN11 Neural Networks Part B © Dr Richard Mitchell 2014 SE2NN11 Neural Networks : Part B In Part B of the course, I Networks Including Stochastic Diffusion Search p2 RJM 16/09/14 SE2NN11 Neural Networks Part B © Dr
Deep Recursive Neural Networks for Compositionality in Language
Cardie, Claire
Deep Recursive Neural Networks for Compositionality in Language Ozan Irsoy Department of Computer representations. Even though these architectures are deep in structure, they lack the capacity for hierarchical representation that exists in conventional deep feed-forward networks as well as in recently investigated deep
Electric Power System Anomaly Detection Using Neural Networks
Tronci, Enrico
to hijacking of measures, changes in the power network topology (i.e. transmission lines breaking) and unexElectric Power System Anomaly Detection Using Neural Networks Marco Martinelli1 , Enrico Tronci1. The aim of this work is to propose an approach to monitor and protect Electric Power System by learning
Neural Network Modeling of Degradation of Solar Cells
Gupta, Himanshu; Ghosh, Bahniman [Department of Electrical Engineering, Indian Institute of Technology, Kanpur, 208016 (India); Banerjee, Sanjay K. [Department of Electrical and Computer Engineering, University of Texas at Austin, Austin, TX, 78758 (United States)
2011-05-25
Neural network modeling has been used to predict the degradation in conversion efficiency of solar cells in this work. The model takes intensity of light, temperature and exposure time as inputs and predicts the conversion efficiency of the solar cell. Backpropagation algorithm has been used to train the network. It is found that the neural network model satisfactorily predicts the degradation in efficiency of the solar cell with exposure time. The error in the computed results, after comparison with experimental results, lies in the range of 0.005-0.01, which is quite low.
Neural networks using two-component Bose-Einstein condensates
Tim Byrnes; Shinsuke Koyama; Kai Yan; Yoshihisa Yamamoto
2012-11-16
The authors previously considered a method solving optimization problems by using a system of interconnected network of two component Bose-Einstein condensates (Byrnes, Yan, Yamamoto New J. Phys. 13, 113025 (2011)). The use of bosonic particles was found to give a reduced time proportional to the number of bosons N for solving Ising model Hamiltonians by taking advantage of enhanced bosonic cooling rates. In this paper we consider the same system in terms of neural networks. We find that up to the accelerated cooling of the bosons the previously proposed system is equivalent to a stochastic continuous Hopfield network. This makes it clear that the BEC network is a physical realization of a simulated annealing algorithm, with an additional speedup due to bosonic enhancement. We discuss the BEC network in terms of typical neural network tasks such as learning and pattern recognition and find that the latter process may be accelerated by a factor of N.
Representing liquid-vapor equilibria of Ternary systems using neural networks
Swisher, Mathew M
2015-01-01
We develop a method based on neural networks for efficiently interpolating equations of state (EOS) for liquid-vapor equilibria of ternary mixtures. We investigate the performance of neural networks both when experimental ...
An adaptive nonlinear MOS scheme for precipitation forecasts using neural networks
Hsieh, William
An adaptive nonlinear MOS scheme for precipitation forecasts using neural networks Yuval, William W A novel neural network (NN) based scheme performs nonlinear Model Output Statistics (MOS) for generating
Senkan, Selim M.
Experimental and artificial neural network modeling study on soot formation in premixed hydrocarbon classical light scattering measurement techniques. The experimental data revealed that the soot properties rights reserved. Keywords: Soot; Hydrocarbon flames; Artificial neural networks 1. Introduction
Use of neural networks to correlate enzymatic hydrolysis with biomass properties
Narayan, Ramasubramanian
2001-01-01
A neural network was used to correlate enzymatic digestibility with the following biomass properties: lignin content, acetyl content, and crystallinity index (CrI). The neural network model was not able to improve a ...
Neural Networks ( ) Contents lists available at SciVerse ScienceDirect
Minai, Ali A.
Neural Networks ( ) Contents lists available at SciVerse ScienceDirect Neural Networks journal is still not completely understood. Even the simplest movements such as reaching for a glass of water
Storage capacity and retrieval time of small-world neural networks
Oshima, Hiraku; Odagaki, Takashi [Department of Physics, Kyushu University, Fukuoka 812-8581 (Japan)
2007-09-15
To understand the influence of structure on the function of neural networks, we study the storage capacity and the retrieval time of Hopfield-type neural networks for four network structures: regular, small world, random networks generated by the Watts-Strogatz (WS) model, and the same network as the neural network of the nematode Caenorhabditis elegans. Using computer simulations, we find that (1) as the randomness of network is increased, its storage capacity is enhanced; (2) the retrieval time of WS networks does not depend on the network structure, but the retrieval time of C. elegans's neural network is longer than that of WS networks; (3) the storage capacity of the C. elegans network is smaller than that of networks generated by the WS model, though the neural network of C. elegans is considered to be a small-world network.
Using a neural network for abnormal event identification in BWRs
Ohga, Yukiharu; Seki, Hiroshi (Hitachi Ltd., Ibaraki (Japan))
1991-01-01
Information on anomalies such as abnormal events is considered to be important for operation support when choosing information to be offered to operators. The authors have applied neural network techniques to identify an abnormal event that causes a reactor scram in boiling water reactors. A primary feature of the method is that the result of the neural network is confirmed using the knowledge base on plant status when each event occurs. This improves the result's reliability. A second feature is that the neural network uses analog data such as reactor pressure, the acquisition of which is triggered by the scram signal. The event identification method is shown. The event identification method is tested using a workstation.
Real-time neural network earthquake profile predictor
Leach, Richard R. (Castro Valley, CA); Dowla, Farid U. (Castro Valley, CA)
1996-01-01
A neural network has been developed that uses first-arrival energy to predict the characteristics of impending earthquake seismograph signals. The propagation of ground motion energy through the earth is a highly nonlinear function. This is due to different forms of ground motion as well as to changes in the elastic properties of the media throughout the propagation path. The neural network is trained using seismogram data from earthquakes. Presented with a previously unseen earthquake, the neural network produces a profile of the complete earthquake signal using data from the first seconds of the signal. This offers a significant advance in the real-time monitoring, warning, and subsequent hazard minimization of catastrophic ground motion.
Real-time neural network earthquake profile predictor
Leach, R.R.; Dowla, F.U.
1996-02-06
A neural network has been developed that uses first-arrival energy to predict the characteristics of impending earthquake seismograph signals. The propagation of ground motion energy through the earth is a highly nonlinear function. This is due to different forms of ground motion as well as to changes in the elastic properties of the media throughout the propagation path. The neural network is trained using seismogram data from earthquakes. Presented with a previously unseen earthquake, the neural network produces a profile of the complete earthquake signal using data from the first seconds of the signal. This offers a significant advance in the real-time monitoring, warning, and subsequent hazard minimization of catastrophic ground motion. 17 figs.
Nuclear power plant fault-diagnosis using artificial neural networks
Kim, Keehoon; Aljundi, T.L.; Bartlett, E.B.
1992-01-01
Artificial neural networks (ANNs) have been applied to various fields due to their fault and noise tolerance and generalization characteristics. As an application to nuclear engineering, we apply neural networks to the early recognition of nuclear power plant operational transients. If a transient or accident occurs, the network will advise the plant operators in a timely manner. More importantly, we investigate the ability of the network to provide a measure of the confidence level in its diagnosis. In this research an ANN is trained to diagnose the status of the San Onofre Nuclear Generation Station using data obtained from the plant's training simulator. Stacked generalization is then applied to predict the error in the ANN diagnosis. The data used consisted of 10 scenarios that include typical design basis accidents as well as less severe transients. The results show that the trained network is capable of diagnosing all 10 instabilities as well as providing a measure of the level of confidence in its diagnoses.
Characteristic functions and process identification by neural networks
Dente, J A
1997-01-01
Principal component analysis (PCA) algorithms use neural networks to extract the eigenvectors of the correlation matrix from the data. However, if the process is non-Gaussian, PCA algorithms or their higher order generalisations provide only incomplete or misleading information on the statistical properties of the data. To handle such situations we propose neural network algorithms, with an hybrid (supervised and unsupervised) learning scheme, which constructs the characteristic function of the probability distribution and the transition functions of the stochastic process. Illustrative examples are presented, which include Cauchy and Levy-type processes
Use of neurals networks in nuclear power plant diagnostics
Uhrig, R.E. (Tennessee Univ., Knoxville, TN (USA). Dept. of Nuclear Engineering Oak Ridge National Lab., TN (USA))
1989-01-01
A technique using neural networks as a means of diagnosing transients or abnormal conditions in nuclear power plants is investigated and found to be feasible. The technique is based on the fact that each physical state of the plant can be represented by a unique pattern of sensor outputs or instrument readings that can be related to the condition of the plant. Neural networks are used to relate this pattern to the fault, problem, or transient condition of the plant. A demonstration of the ability of this technique to identify causes of perturbations in the steam generator of a nuclear plant is presented. 3 refs., 4 figs.
On the Rejection of Fake Muons in AMANDA using Neural Networks
Wiebusch, Christopher
On the Rejection of Fake Muons in AMANDA using Neural Networks Alexander Biron, Sabine Schilling This report describes the use of an artificial neural network in the final quality analysis for AMANDA. Aim the reconstructed tracks and are fed as inputs into the neural network. We use a simple feedforward architecture
Neural Networks Approaches for Discovering the Learnable Correlation between Gene Function and Gene
Bonner, Anthony
Neural Networks Approaches for Discovering the Learnable Correlation between Gene Function and Gene novel clustering and Neural Network (NN) approaches for predicting mouse gene functions from gene. Our results show that neural networks can be extremely useful in this area. We present the improved
Prediction of Protein Functional Domains from Sequences Using Artificial Neural Networks
Szepesvari, Csaba
Prediction of Protein Functional Domains from Sequences Using Artificial Neural Networks János Ltd., Budapest 1121, Hungary An artificial neural network (ANN) solution is described, are forwarded to feed-forward artificial neural networks with six input and six hidden units with sigmoidal
LETTER Communicated by Daniel Amit Effects of Fast Presynaptic Noise in Attractor Neural Networks
Garrido, Pedro L.
LETTER Communicated by Daniel Amit Effects of Fast Presynaptic Noise in Attractor Neural Networks J neural networks. The noise occurs on a very short timescale compared to that for the neuron dynamics. We thus describe a mechanism by which fast presynaptic noise enhances the neural network sensitivity
Pruned neural networks for regression Rudy Setiono and Wee Kheng Leow
Leow, Wee Kheng
Pruned neural networks for regression Rudy Setiono and Wee Kheng Leow School of Computing National University of Singapore Singapore 117543 Abstract. Neural networks have been widely used as a tool for regres hand, using the mean absolute error as the measurement metric, the neural network method outperforms
Neural networks as a tool for constructing continuous NDVI time series from AVHRR and MODIS
Neural networks as a tool for constructing continuous NDVI time series from AVHRR and MODIS M. E-stationary aspect of NDVI, we use an artificial neural network (ANN) to map the NDVI indices from AVHRR to those the AVHRR record. Four years of overlap between the two sensors is used to train a neural network to remove
Contributed article Fractional Fourier transform pre-processing for neural networks
Barshan, Billur
Contributed article Fractional Fourier transform pre-processing for neural networks and its This study investigates fractional Fourier transform pre-processing of input signals to neural networks. Judicious choice of this parameter can lead to overall improvement of the neural network performance
Neural Networks and Information in Materials Science H. K. D. H. Bhadeshia*
Cambridge, University of
REVIEW Neural Networks and Information in Materials Science H. K. D. H. Bhadeshia* Department in Wiley InterScience (www.interscience.wiley.com). Abstract: Neural networks have pervaded all aspects 1. INTRODUCTION Neural networks are wonderful tools, which permit the development of quantitative
Neural-Network-Based Fuzzy Identifier: Design and Evaluation Dr. Kasim M. Al-Aubidy
Neural-Network-Based Fuzzy Identifier: Design and Evaluation Dr. Kasim M. Al-Aubidy Computer is represented as a feed forward neural network, is able to incorporate qualitative and quantitative information into adaptive fuzzy and neural networks[3]. The proposed approach handles an architecture which is somehow
Automatic Language Identification using Long Short-Term Memory Recurrent Neural Networks
Cortes, Corinna
Automatic Language Identification using Long Short-Term Memory Recurrent Neural Networks Javier-Term Memory (LSTM) recurrent neural networks (RNNs) for automatic lan- guage identification (LID). The use is compared to baseline i-vector and feed forward Deep Neural Network (DNN) systems in the NIST Language
Automatic Assessment of Levodopa-Induced Dyskinesias in Daily Life by Neural Networks
Gielen, C.C.A.M.
Automatic Assessment of Levodopa-Induced Dyskinesias in Daily Life by Neural Networks NoeÂ¨l L on the body. A neural network was trained to assess the severity of LID using various variables of the accelerometer signals. Neural network scores were compared with the assessment by physicians, who evaluated
Z .Applied Surface Science 149 1999 97102 Unfolding positron lifetime spectra with neural networks
Pázsit, Imre
Z .Applied Surface Science 149 1999 97102 Unfolding positron lifetime spectra with neural networks is based on the use of artificial neural networks ANNs . By using data from simulated positron spectra: Artificial neural networks ANNs ; Amplitudes; Simulation model 1. Introduction Determination of mean
Pilyugin, Sergei S.
716 IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 14, NO. 3, MAY 2003 Letters__________________________________________________________________________________________ Global Exponential Stability of Competitive Neural Networks With Different Time Scales A. Meyer-Baese, S- organization must include the aspects of long and short-term memory. The behavior of such a neural network
Neural Networks, Vol. 5, pp. 735-743, 1992 Printed in th USA. Ail rights reserved.
Chapeau-Blondeau, François
Neural Networks, Vol. 5, pp. 735-743, 1992 Printed in thé USA. Ail rights reserved. 0893-6080/92 $5 Régimes in thé Dynamics of Small Neural Networks With Delay FRANÇOIS CHAPEAU-BLONDEAU AND GILBERT CHAUVET considersimple neural network models consisting oftwo to three continuonsnonlinear neurons, with no intrinsic
Batch-mode vs Online-mode Supervised Learning Motivations for Artificial Neural Networks
Wehenkel, Louis
Batch-mode vs Online-mode Supervised Learning Motivations for Artificial Neural Networks Linear ANN-mode vs Online-mode Supervised Learning Motivations for Artificial Neural Networks Linear ANN Models for Artificial Neural Networks Linear ANN Models Single neuron models Single layer models Nonlinear ANN Models
Communicated by Rodney Brooks A Distributed Neural Network Architecture for Hexapod
Beer, Randall D.
Communicated by Rodney Brooks A Distributed Neural Network Architecture for Hexapod Robot present a fully distributed neural network architecture for control- ling the locomotion of a hexapod of Technology #12;Neural Network Architecturefor Hexapod Robot Locomotion 357 Figure 1: A comparison
Neural NILM: Deep Neural Networks Applied to Energy Disaggregation
Knottenbelt, William J.
of Computing Imperial College London 180 Queen's Gate, London, SW7 2RH, UK jack.kelly@imperial.ac.uk William Knottenbelt Department of Computing Imperial College London 180 Queen's Gate, London, SW7 2RH, UK w) denoising autoencoders; and 3) a network which regresses the start time, end time and average power demand
Mathematically Reduced Chemical Reaction Mechanism Using Neural Networks
Nelson Butuk
2004-12-01
This is an annual technical report for the work done over the last year (period ending 9/30/2004) on the project titled ''Mathematically Reduced Chemical Reaction Mechanism Using Neural Networks''. The aim of the project is to develop an efficient chemistry model for combustion simulations. The reduced chemistry model will be developed mathematically without the need of having extensive knowledge of the chemistry involved. To aid in the development of the model, Neural Networks (NN) will be used via a new network topology know as Non-linear Principal Components Analysis (NPCA). We report on the development of a procedure to speed up the training of NPCA. The developed procedure is based on the non-parametric statistical technique of kernel smoothing. When this smoothing technique is implemented as a Neural Network, It is know as Generalized Regression Neural Network (GRNN). We present results of implementing GRNN on a test problem. In addition, we present results of an in house developed 2-D CFD code that will be used through out the project period.
Mathematically Reduced Chemical Reaction Mechanism Using Neural Networks
Ziaul Huque
2007-08-31
This is the final technical report for the project titled 'Mathematically Reduced Chemical Reaction Mechanism Using Neural Networks'. The aim of the project was to develop an efficient chemistry model for combustion simulations. The reduced chemistry model was developed mathematically without the need of having extensive knowledge of the chemistry involved. To aid in the development of the model, Neural Networks (NN) was used via a new network topology known as Non-linear Principal Components Analysis (NPCA). A commonly used Multilayer Perceptron Neural Network (MLP-NN) was modified to implement NPCA-NN. The training rate of NPCA-NN was improved with the GEneralized Regression Neural Network (GRNN) based on kernel smoothing techniques. Kernel smoothing provides a simple way of finding structure in data set without the imposition of a parametric model. The trajectory data of the reaction mechanism was generated based on the optimization techniques of genetic algorithm (GA). The NPCA-NN algorithm was then used for the reduction of Dimethyl Ether (DME) mechanism. DME is a recently discovered fuel made from natural gas, (and other feedstock such as coal, biomass, and urban wastes) which can be used in compression ignition engines as a substitute for diesel. An in-house two-dimensional Computational Fluid Dynamics (CFD) code was developed based on Meshfree technique and time marching solution algorithm. The project also provided valuable research experience to two graduate students.
Stimulus-dependent suppression of chaos in recurrent neural networks
Rajan, Kanaka; Abbott, L. F.; Sompolinsky, Haim [Lewis-Sigler Institute for Integrative Genomics, Icahn 262, Princeton University, Princeton, New Jersey 08544 (United States); Department of Neuroscience and Department of Physiology and Cellular Biophysics, College of Physicians and Surgeons, Columbia University, New York, New York 10032-2695 (United States); Racah Institute of Physics, Interdisciplinary Center for Neural Computation, Hebrew University, Jerusalem (Israel)
2010-07-15
Neuronal activity arises from an interaction between ongoing firing generated spontaneously by neural circuits and responses driven by external stimuli. Using mean-field analysis, we ask how a neural network that intrinsically generates chaotic patterns of activity can remain sensitive to extrinsic input. We find that inputs not only drive network responses, but they also actively suppress ongoing activity, ultimately leading to a phase transition in which chaos is completely eliminated. The critical input intensity at the phase transition is a nonmonotonic function of stimulus frequency, revealing a 'resonant' frequency at which the input is most effective at suppressing chaos even though the power spectrum of the spontaneous activity peaks at zero and falls exponentially. A prediction of our analysis is that the variance of neural responses should be most strongly suppressed at frequencies matching the range over which many sensory systems operate.
Neural network determination of parton distributions: the nonsinglet case
The NNPDF Collaboration; Luigi Del Debbio; Stefano Forte; Jose I. Latorre; Andrea Piccione; Joan Rojo
2007-01-16
We provide a determination of the isotriplet quark distribution from available deep--inelastic data using neural networks. We give a general introduction to the neural network approach to parton distributions, which provides a solution to the problem of constructing a faithful and unbiased probability distribution of parton densities based on available experimental information. We discuss in detail the techniques which are necessary in order to construct a Monte Carlo representation of the data, to construct and evolve neural parton distributions, and to train them in such a way that the correct statistical features of the data are reproduced. We present the results of the application of this method to the determination of the nonsinglet quark distribution up to next--to--next--to--leading order, and compare them with those obtained using other approaches.
Deep Learning & Neural Networks Graduate School of Information Science
Duh, Kevin
Deep Learning & Neural Networks Lecture 2 Kevin Duh Graduate School of Information Science Nara Institute of Science and Technology Jan 16, 2014 #12;Today's Topics 1 General Ideas in Deep Learning Motivation for Deep Architectures and why is it hard? Main Breakthrough in 2006: Layer-wise Pre-Training 2
Deep Learning & Neural Networks Graduate School of Information Science
Duh, Kevin
Deep Learning & Neural Networks Lecture 1 Kevin Duh Graduate School of Information Science Nara Institute of Science and Technology Jan 14, 2014 #12;2/40 #12;3/40 #12;4/40 #12;What is Deep Learning? A family of methods that uses deep architectures to learn high-level feature representations 5/40 #12;What
A STATISTICAL NEURAL NETWORK FRAMEWORK FOR RISK MANAGEMENT PROCESS
Basili, Victor R.
A STATISTICAL NEURAL NETWORK FRAMEWORK FOR RISK MANAGEMENT PROCESS From the Proposal to its of Maryland, A.V. Williams Bldg. 115, College Park 20742, MD, USA basili@cs.umd.edu Keywords: Risk Management available formal risk management models and related frameworks by providing an independent mechanism
Autonomous Perceptron Neural Network Inspired from Quantum computing
M. Zidan; A. Sagheer; N. Metwally
2015-10-02
Recently with the rapid development of technology, there are a lot of applications require to achieve low-cost learning in order to accomplish inexpensive computation. However the known computational power of classical artificial neural networks (CANN), they are not capable to provide low-cost learning due to many reasons such as linearity, complexity of architecture, etc. In contrast, quantum neural networks (QNN) may be representing a good computational alternate to CANN, based on the computational power of quantum bit (qubit) over the classical bit. In this paper, a new algorithm of quantum perceptron neural network based only on one neuron is introduced to overcome some limitations of the classical perceptron neural networks. The proposed algorithm is capable to construct its own set of activation operators that enough to accomplish the learning process in a limited number of iterations and, consequently, reduces the cost of computation. For evaluation purpose, we utilize the proposed algorithm to solve five problems using real and artificial data. It is shown throughout the paper that promising results are provided and compared favorably with other reported algorithms
Use of Artificial Neural Networks Process Analyzers: A Case Study
Ghouti, Lahouari
for this case study was obtained from SHARQ petrochemical company, Saudi Arabia. Four process variablesUse of Artificial Neural Networks Process Analyzers: A Case Study Al-Duwaish1 , H., Ghouti2 , L to predict O2 contents in a boiler at SHARQ petrochemical company in Saudi Arabia. The training data has been
Language discrimination and clustering via a neural network approach
Mariano, Angelo; Pascazio, Saverio
2015-01-01
We classify twenty-one Indo-European languages starting from written text. We use neural networks in order to define a distance among different languages, construct a dendrogram and analyze the ultrametric structure that emerges. Four or five subgroups of languages are identified, according to the "cut" of the dendrogram, drawn with an entropic criterion. The results and the method are discussed.
Studying Deeply Virtual Compton Scattering with Neural Networks
Kresimir Kumericki; Dieter Mueller; Andreas Schafer
2011-10-17
Neural networks are utilized to fit Compton form factor H to HERMES data on deeply virtual Compton scattering off unpolarized protons. We used this result to predict the beam charge-spin assymetry for muon scattering off proton at the kinematics of the COMPASS II experiment.
Evolving Plastic Neural Networks with Novelty Search Sebastian Risi
Stanley, Kenneth O.
Evolving Plastic Neural Networks with Novelty Search Sebastian Risi Charles E. Hughes Kenneth O-based learning tasks, thereby opening a new direction for research in evolving plastic ANNs. Keywords Novelty combining these complementary forms of adaptation by evolving ANNs with synaptic plastic- ity driven
Metal Oxide Semiconductor Gas Sensors and Neural Networks
Siegel, Mel
Olfaction Metal Oxide Semiconductor Gas Sensors and Neural Networks M. W. Siegel Carnegie Mellon around a chemical plant, sniffing as it goes for gas leaks (or the vapors of liquid leaks), navigating perhaps directed to the offending pipe fissure or open valve by acoustic homing toward the source
Hybrid coupled modeling of the tropical Pacific using neural networks
Hsieh, William
Hybrid coupled modeling of the tropical Pacific using neural networks Shuyong Li, William W. Hsieh To investigate the potential for improving hybrid coupled models (HCM) of the tropical Pacific by the use: dynamical coupled models, statistical models and hybrid coupled models [Barnston et al., 1994]. A hybrid
Neural Networks for Post-processing Model Output: Caren Marzban
Marzban, Caren
variables to the neural network are: Forecast hour, model forecast temperature, relative humidity, wind direction and speed, mean sea level pressure, cloud cover, and precipitation rate and amount. The single to being able to approximate a large class of functions, they are less inclined to overfit data than some
Statistical Mechanics of Recurrent Neural Networks I Statics
Coolen, ACC "Ton"
CHAPTER 14 Statistical Mechanics of Recurrent Neural Networks I Â± Statics A.C.C. COOLEN Department . . . . . . . . . . . . . . . . . . . . . . . . . . 540 2.3. Detailed balance and equilibrium statistical mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595 532 #12;1. Introduction Statistical mechanics deals with large systems of stochastically
A Neural Network-based ARX Model of Virgo Noise
F. Barone; R. De Rosa; A. Eleuteri; F. Garufi; L. Milano; R. Tagliaferri
1999-06-07
In this paper a Neural Network based approach is presented to identify the noise in the VIRGO context. VIRGO is an experiment to detect Gravitational Waves by means of a Laser Interferometer. Preliminary results appear to be very promising for data analysis of realistic Interferometer outputs.
NONLINEAR MULTIVARIATE AND TIME SERIES ANALYSIS BY NEURAL NETWORK METHODS
Hsieh, William
NONLINEAR MULTIVARIATE AND TIME SERIES ANALYSIS BY NEURAL NETWORK METHODS William W. Hsieh] Methods in multivariate statistical analysis are essential for working with large amounts of geophysical multivariate statistical analysis, there is a hierarchy of methods, starting with linear regression at the base
Wind Power Plant Prediction by Using Neural Networks: Preprint
Liu, Z.; Gao, W.; Wan, Y. H.; Muljadi, E.
2012-08-01
This paper introduces a method of short-term wind power prediction for a wind power plant by training neural networks based on historical data of wind speed and wind direction. The model proposed is shown to achieve a high accuracy with respect to the measured data.
Forecasting Hourly Electricity Load Profile Using Neural Networks
Koprinska, Irena
Forecasting Hourly Electricity Load Profile Using Neural Networks Mashud Rana and Irena Koprinska--We present INN, a new approach for predicting the hourly electricity load profile for the next day from a time series of previous electricity loads. It uses an iterative methodology to make the predictions
Vibration monitoring of EDF rotating machinery using artificial neural networks
Alguindigue, I.E.; Loskiewicz-Buczak, A.; Uhrig, R.E. (Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering); Hamon, L.; Lefevre, F. (Electricite de France, 78 - Chatou (France). Direction des Etudes et Recherches)
1991-01-01
Vibration monitoring of components in nuclear power plants has been used for a number of years. This technique involves the analysis of vibration data coming from vital components of the plant to detect features which reflect the operational state of machinery. The analysis leads to the identification of potential failures and their causes, and makes it possible to perform efficient preventive maintenance. Earlydetection is important because it can decrease the probability of catastrophic failures, reduce forced outgage, maximize utilization of available assets, increase the life of the plant, and reduce maintenance costs. This paper documents our work on the design of a vibration monitoring methodology based on neural network technology. This technology provides an attractive complement to traditional vibration analysis because of the potential of neural networks to operate in real-time mode and to handle data which may be distorted or noisy. Our efforts have been concentrated on the analysis and classification of vibration signatures collected by Electricite de France (EDF). Two neural networks algorithms were used in our project: the Recirculation algorithm and the Backpropagation algorithm. Although this project is in the early stages of development it indicates that neural networks may provide a viable methodology for monitoring and diagnostics of vibrating components. Our results are very encouraging.
Vibration monitoring of EDF rotating machinery using artificial neural networks
Alguindigue, I.E.; Loskiewicz-Buczak, A.; Uhrig, R.E. [Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering; Hamon, L.; Lefevre, F. [Electricite de France, 78 - Chatou (France). Direction des Etudes et Recherches
1991-12-31
Vibration monitoring of components in nuclear power plants has been used for a number of years. This technique involves the analysis of vibration data coming from vital components of the plant to detect features which reflect the operational state of machinery. The analysis leads to the identification of potential failures and their causes, and makes it possible to perform efficient preventive maintenance. Earlydetection is important because it can decrease the probability of catastrophic failures, reduce forced outgage, maximize utilization of available assets, increase the life of the plant, and reduce maintenance costs. This paper documents our work on the design of a vibration monitoring methodology based on neural network technology. This technology provides an attractive complement to traditional vibration analysis because of the potential of neural networks to operate in real-time mode and to handle data which may be distorted or noisy. Our efforts have been concentrated on the analysis and classification of vibration signatures collected by Electricite de France (EDF). Two neural networks algorithms were used in our project: the Recirculation algorithm and the Backpropagation algorithm. Although this project is in the early stages of development it indicates that neural networks may provide a viable methodology for monitoring and diagnostics of vibrating components. Our results are very encouraging.
Proceedings of the Neural Network Workshop for the Hanford Community
Keller, P.E.
1994-01-01
These proceedings were generated from a series of presentations made at the Neural Network Workshop for the Hanford Community. The abstracts and viewgraphs of each presentation are reproduced in these proceedings. This workshop was sponsored by the Computing and Information Sciences Department in the Molecular Science Research Center (MSRC) at the Pacific Northwest Laboratory (PNL). Artificial neural networks constitute a new information processing technology that is destined within the next few years, to provide the world with a vast array of new products. A major reason for this is that artificial neural networks are able to provide solutions to a wide variety of complex problems in a much simpler fashion than is possible using existing techniques. In recognition of these capabilities, many scientists and engineers are exploring the potential application of this new technology to their fields of study. An artificial neural network (ANN) can be a software simulation, an electronic circuit, optical system, or even an electro-chemical system designed to emulate some of the brain`s rudimentary structure as well as some of the learning processes that are believed to take place in the brain. For a very wide range of applications in science, engineering, and information technology, ANNs offer a complementary and potentially superior approach to that provided by conventional computing and conventional artificial intelligence. This is because, unlike conventional computers, which have to be programmed, ANNs essentially learn from experience and can be trained in a straightforward fashion to carry out tasks ranging from the simple to the highly complex.
Forecasting Hospital Bed Availability Using Simulation and Neural Networks
Kuhl, Michael E.
Forecasting Hospital Bed Availability Using Simulation and Neural Networks Matthew J. Daniels, NY 14623 Elisabeth Hager Hager Consulting Pittsford, NY 14534 Abstract The availability of beds is a critical factor for decision-making in hospitals. Bed availability (or alternatively the bed occupancy
Introduction to Neural Networks : Important Equations John A. Bullinaria -2004
Bullinaria, John
out f w inj ji i i n j= - = ( ) 1 Sigmoid/Logistic activation function Sigmoid( )x e x= + - 1 1 Sum in w inkl l l n n nl k p = -( ) . ( ). Back-propagation with momentum delta delta w f out wk n k n lk n k j n jk n j ( ) ( ) ( ) ( ) ( ) . .= + + - 1 1 1 w t delta t out t w tkl n l n p k
Franklin, James
How a neural net grows symbols Proceedings of the Seventh Australian Conference on Neural Networks neural networks and symbolic AI, in such a way as to combine the good features of each. It is argued.Franklin@unsw.edu.au April 8, 2005 Abstract Brains, unlike artificial neural nets, use sym- bols to summarise and reason
Enhanced memory performance thanks to neural network assortativity
Franciscis, S. de; Johnson, S.; Torres, J. J. [Departamento de Electromagnetismo y Fisica de la Materia, and Institute Carlos I for Theoretical and Computational Physics, Facultad de Ciencias, University of Granada, 18071 Granada (Spain)
2011-03-24
The behaviour of many complex dynamical systems has been found to depend crucially on the structure of the underlying networks of interactions. An intriguing feature of empirical networks is their assortativity--i.e., the extent to which the degrees of neighbouring nodes are correlated. However, until very recently it was difficult to take this property into account analytically, most work being exclusively numerical. We get round this problem by considering ensembles of equally correlated graphs and apply this novel technique to the case of attractor neural networks. Assortativity turns out to be a key feature for memory performance in these systems - so much so that for sufficiently correlated topologies the critical temperature diverges. We predict that artificial and biological neural systems could significantly enhance their robustness to noise by developing positive correlations.
A stochastic learning algorithm for layered neural networks
Bartlett, E.B. (Iowa State Univ. of Science and Technology, Ames, IA (United States). Dept. of Mechanical Engineering); Uhrig, R.E. (Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering)
1992-01-01
The random optimization method typically uses a Gaussian probability density function (PDF) to generate a random search vector. In this paper the random search technique is applied to the neural network training problem and is modified to dynamically seek out the optimal probability density function (OPDF) from which to select the search vector. The dynamic OPDF search process, combined with an auto-adaptive stratified sampling technique and a dynamic node architecture (DNA) learning scheme, completes the modifications of the basic method. The DNA technique determines the appropriate number of hidden nodes needed for a given training problem. By using DNA, researchers do not have to set the neural network architectures before training is initiated. The approach is applied to networks of generalized, fully interconnected, continuous perceptions. Computer simulation results are given.
A stochastic learning algorithm for layered neural networks
Bartlett, E.B. [Iowa State Univ. of Science and Technology, Ames, IA (United States). Dept. of Mechanical Engineering] [Iowa State Univ. of Science and Technology, Ames, IA (United States). Dept. of Mechanical Engineering; Uhrig, R.E. [Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering] [Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering
1992-12-31
The random optimization method typically uses a Gaussian probability density function (PDF) to generate a random search vector. In this paper the random search technique is applied to the neural network training problem and is modified to dynamically seek out the optimal probability density function (OPDF) from which to select the search vector. The dynamic OPDF search process, combined with an auto-adaptive stratified sampling technique and a dynamic node architecture (DNA) learning scheme, completes the modifications of the basic method. The DNA technique determines the appropriate number of hidden nodes needed for a given training problem. By using DNA, researchers do not have to set the neural network architectures before training is initiated. The approach is applied to networks of generalized, fully interconnected, continuous perceptions. Computer simulation results are given.
DEEP NEURAL NETWORKS FOR SMALL FOOTPRINT TEXT-DEPENDENT SPEAKER VERIFICATION
Cortes, Corinna
DEEP NEURAL NETWORKS FOR SMALL FOOTPRINT TEXT-DEPENDENT SPEAKER VERIFICATION Ehsan Variani1 , Xin the use of deep neural networks (DNNs) for a small footprint text-dependent speaker verification task% relative in equal error rate (EER) for clean and noisy conditions respectively. Index Terms-- Deep neural
Neural network uncertainty assessment using Bayesian statistics with application to remote sensing
Aires, Filipe
Neural network uncertainty assessment using Bayesian statistics with application to remote sensing for many inversion problems in remote sensing; however, uncertainty estimates are rarely provided Meteorology and Atmospheric Dynamics: General or miscellaneous; KEYWORDS: remote sensing, uncertainty, neural
Scalable network-on-chip architecture for configurable neural networks Dmitri Vainbrand
Ginosar, Ran
implementation Interconnect architecture a b s t r a c t Providing highly flexible connectivity is a major. The high level architectural framework for this research is a multiprocessing chip (CMP) comprising a largeScalable network-on-chip architecture for configurable neural networks Dmitri Vainbrand , Ran
Measuring photometric redshifts using galaxy images and Deep Neural Networks
Hoyle, Ben
2015-01-01
We propose a new method to estimate the photometric redshift of galaxies by using the full galaxy image in each measured band. This method draws from the latest techniques and advances in machine learning, in particular Deep Neural Networks. We pass the entire multi-band galaxy image into the machine learning architecture to obtain a redshift estimate that is competitive with the best existing standard machine learning techniques. The standard techniques estimate redshifts using post-processed features, such as magnitudes and colours, which are extracted from the galaxy images and are deemed to be salient by the user. This new method removes the user from the photometric redshift estimation pipeline. However we do note that Deep Neural Networks require many orders of magnitude more computing resources than standard machine learning architectures.
Adaptive model predictive process control using neural networks
Buescher, Kevin L. (Los Alamos, NM); Baum, Christopher C. (Mazomanie, WI); Jones, Roger D. (Espanola, NM)
1997-01-01
A control system for controlling the output of at least one plant process output parameter is implemented by adaptive model predictive control using a neural network. An improved method and apparatus provides for sampling plant output and control input at a first sampling rate to provide control inputs at the fast rate. The MPC system is, however, provided with a network state vector that is constructed at a second, slower rate so that the input control values used by the MPC system are averaged over a gapped time period. Another improvement is a provision for on-line training that may include difference training, curvature training, and basis center adjustment to maintain the weights and basis centers of the neural in an updated state that can follow changes in the plant operation apart from initial off-line training data.
Adaptive model predictive process control using neural networks
Buescher, K.L.; Baum, C.C.; Jones, R.D.
1997-08-19
A control system for controlling the output of at least one plant process output parameter is implemented by adaptive model predictive control using a neural network. An improved method and apparatus provides for sampling plant output and control input at a first sampling rate to provide control inputs at the fast rate. The MPC system is, however, provided with a network state vector that is constructed at a second, slower rate so that the input control values used by the MPC system are averaged over a gapped time period. Another improvement is a provision for on-line training that may include difference training, curvature training, and basis center adjustment to maintain the weights and basis centers of the neural in an updated state that can follow changes in the plant operation apart from initial off-line training data. 46 figs.
Neural network based design of cellular manufacturing systems
Ramachandran, Satheesh
1990-01-01
, India Chair of Advisory Committee: Dr. Cesar O. Malave A neural network based on a competitive learning rule when trained with the part machine incidence matrix of a large number of parts classifies the parts and machines into part families... and machine cells respectively. This classification compares well with the classical clustering techniques. The steady state values of the activations and interconnecting strengths enables easier identification of the part families, machine cells...
Process for forming synapses in neural networks and resistor therefor
Fu, Chi Y. (San Francisco, CA)
1996-01-01
Customizable neural network in which one or more resistors form each synapse. All the resistors in the synaptic array are identical, thus simplifying the processing issues. Highly doped, amorphous silicon is used as the resistor material, to create extremely high resistances occupying very small spaces. Connected in series with each resistor in the array is at least one severable conductor whose uppermost layer has a lower reflectivity of laser energy than typical metal conductors at a desired laser wavelength.
Laser programmable integrated curcuit for forming synapses in neural networks
Fu, Chi Y. (San Francisco, CA)
1997-01-01
Customizable neural network in which one or more resistors form each synapse. All the resistors in the synaptic array are identical, thus simplifying the processing issues. Highly doped, amorphous silicon is used as the resistor material, to create extremely high resistances occupying very small spaces. Connected in series with each resistor in the array is at least one severable conductor whose uppermost layer has a lower reflectivity of laser energy than typical metal conductors at a desired laser wavelength.
Re-evaluation of the Gottfried sum using neural networks
Riccardo Abbate; Stefano Forte
2005-11-18
We provide a determination of the Gottfried sum from all available data, based on a neural network parametrization of the nonsinglet structure function F_2. We find S_G=0.244 +- 0.045, closer to the quark model expectation S_G=1/3 than previous results. We show that the uncertainty from the small x region is somewhat underestimated in previous determinations.
Automatic classification of eclipsing binaries light curves using neural networks
L. M. Sarro; C. Sánchez-Fernández; A. Giménez
2005-11-11
In this work we present a system for the automatic classification of the light curves of eclipsing binaries. This system is based on a classification scheme that aims to separate eclipsing binary sistems according to their geometrical configuration in a modified version of the traditional classification scheme. The classification is performed by a Bayesian ensemble of neural networks trained with {\\em Hipparcos} data of seven different categories including eccentric binary systems and two types of pulsating light curve morphologies.
Laser programmable integrated circuit for forming synapses in neural networks
Fu, C.Y.
1997-02-11
Customizable neural network in which one or more resistors form each synapse is disclosed. All the resistors in the synaptic array are identical, thus simplifying the processing issues. Highly doped, amorphous silicon is used as the resistor material, to create extremely high resistances occupying very small spaces. Connected in series with each resistor in the array is at least one severable conductor whose uppermost layer has a lower reflectivity of laser energy than typical metal conductors at a desired laser wavelength. 5 figs.
Process for forming synapses in neural networks and resistor therefor
Fu, C.Y.
1996-07-23
Customizable neural network in which one or more resistors form each synapse is disclosed. All the resistors in the synaptic array are identical, thus simplifying the processing issues. Highly doped, amorphous silicon is used as the resistor material, to create extremely high resistances occupying very small spaces. Connected in series with each resistor in the array is at least one severable conductor whose uppermost layer has a lower reflectivity of laser energy than typical metal conductors at a desired laser wavelength. 5 figs.
Nuclear power plant fault-diagnosis using artificial neural networks
Kim, Keehoon; Aljundi, T.L.; Bartlett, E.B.
1992-12-31
Artificial neural networks (ANNs) have been applied to various fields due to their fault and noise tolerance and generalization characteristics. As an application to nuclear engineering, we apply neural networks to the early recognition of nuclear power plant operational transients. If a transient or accident occurs, the network will advise the plant operators in a timely manner. More importantly, we investigate the ability of the network to provide a measure of the confidence level in its diagnosis. In this research an ANN is trained to diagnose the status of the San Onofre Nuclear Generation Station using data obtained from the plant`s training simulator. Stacked generalization is then applied to predict the error in the ANN diagnosis. The data used consisted of 10 scenarios that include typical design basis accidents as well as less severe transients. The results show that the trained network is capable of diagnosing all 10 instabilities as well as providing a measure of the level of confidence in its diagnoses.
Saini, K. K.; Saini, Sanju [CDLM engg. College Panniwala Mota, Sirsa and Murthal, Sonipat, Haryana (India)
2008-10-07
Neural networks are a relatively new artificial intelligence technique that emulates the behavior of biological neural systems in digital software or hardware. These networks can 'learn', automatically, complex relationships among data. This feature makes the technique very useful in modeling processes for which mathematical modeling is difficult or impossible. The work described here outlines some examples of the application of neural networks with artificial intelligence technique in the modeling of industrial processes.
Topological and Dynamical Complexity of Random Neural Networks
Gilles Wainrib; Jonathan Touboul
2013-03-15
Random neural networks are dynamical descriptions of randomly interconnected neural units. These show a phase transition to chaos as a disorder parameter is increased. The microscopic mechanisms underlying this phase transition are unknown, and similarly to spin-glasses, shall be fundamentally related to the behavior of the system. In this Letter we investigate the explosion of complexity arising near that phase transition. We show that the mean number of equilibria undergoes a sharp transition from one equilibrium to a very large number scaling exponentially with the dimension on the system. Near criticality, we compute the exponential rate of divergence, called topological complexity. Strikingly, we show that it behaves exactly as the maximal Lyapunov exponent, a classical measure of dynamical complexity. This relationship unravels a microscopic mechanism leading to chaos which we further demonstrate on a simpler class of disordered systems, suggesting a deep and underexplored link between topological and dynamical complexity.
Bump formation in a binary attractor neural network
Koroutchev, Kostadin; Korutcheva, Elka
2006-02-15
The conditions for the formation of local bumps in the activity of binary attractor neural networks with spatially dependent connectivity are investigated. We show that these formations are observed when asymmetry between the activity during the retrieval and learning is imposed. An analytical approximation for the order parameters is derived. The corresponding phase diagram shows a relatively large and stable region where this effect is observed, although critical storage and information capacities drastically decrease inside that region. We demonstrate that the stability of the network, when starting from the bump formation, is larger than the stability when starting even from the whole pattern. Finally, we show a very good agreement between the analytical results and the simulations performed for different topologies of the network.
Pattern classification and associative recall by neural networks
Chiueh, Tzi-Dar.
1989-01-01
The first part of this dissertation discusses a new classifier based on a multilayer feed-forward network architecture. The main idea is to map irregularly-distributed prototypes in a classification problem to codewords that are organized in some way. Then the pattern classification problem is transformed into a threshold decoding problem, which is easily solved using simple hard-limiter neurons. At first the author proposes the new model and introduce two families of good internal representation codes. Then some analyses and software simulation concerning the storage capacity of this new model are done. The results show that the new classifier is much better than the classifier based on the Hopfield model in terms of both the storage capacity and the ability to classify correlated prototypes. A general model for neural network associative memories with a feedback structure is proposed. Many existing neural network associative memories can be expressed as special cases of this general model. Among these models, there is a class of associative memories, called correlation associative memories, that are capable of storing a large number of memory patterns. If the function used in the evolution equation is monotonically nondecreasing, then a correlation associative memory can be proved to be asymptotically stable in both the synchronous and asynchronous updating modes. Of these correlation associative memories, one stands out because of its VLSI implementation feasibility and large storage capacity. This memory uses the exponentiation function in its evolution equation; hence it is called exponential correlation associative memory (ECAM).
Neural network detected in a presumed vestigial trait: ultrastructure of the
Reimchen, Thomas E.
Neural network detected in a presumed vestigial trait: ultrastructure of the salmonid adipose fin J; nerve network; astrocytes; primary cilia; fisheries 1. INTRODUCTION The long-term persistence of traits
Wiggins, Vince L.
1996-01-01
recently demonstrated capabilities in areas important to personnel research such as statistical analysis, decision modeling, control, and forecasting. An extensive review of the neural network literature indicates that these networks have proven superior...
Neural Network Based Montioring and Control of Fluidized Bed.
Bodruzzaman, M.; Essawy, M.A.
1996-04-01
The goal of this project was to develop chaos analysis and neural network-based modeling techniques and apply them to the pressure-drop data obtained from the Fluid Bed Combustion (FBC) system (a small scale prototype model) located at the Federal Energy Technology Center (FETC)-Morgantown. The second goal was to develop neural network-based chaos control techniques and provide a suggestive prototype for possible real-time application to the FBC system. The experimental pressure data were collected from a cold FBC experimental set-up at the Morgantown Center. We have performed several analysis on these data in order to unveil their dynamical and chaotic characteristics. The phase-space attractors were constructed from the one dimensional time series data, using the time-delay embedding method, for both normal and abnormal conditions. Several identifying parameters were also computed from these attractors such as the correlation dimension, the Kolmogorov entropy, and the Lyapunov exponents. These chaotic attractor parameters can be used to discriminate between the normal and abnormal operating conditions of the FBC system. It was found that, the abnormal data has higher correlation dimension, larger Kolmogorov entropy and larger positive Lyapunov exponents as compared to the normal data. Chaotic system control using neural network based techniques were also investigated and compared to conventional chaotic system control techniques. Both types of chaotic system control techniques were applied to some typical chaotic systems such as the logistic, the Henon, and the Lorenz systems. A prototype model for real-time implementation of these techniques has been suggested to control the FBC system. These models can be implemented for real-time control in a next phase of the project after obtaining further measurements from the experimental model. After testing the control algorithms developed for the FBC model, the next step is to implement them on hardware and link them to the experimental system. In this report, the hardware implementation issues of the control algorithms are also discussed.
Neural network and area method interpretation of pulsed experiments
Dulla, S.; Picca, P.; Ravetto, P. [Politecnico di Torino, Dipartimento di Energetica, Corso Duca degli Abruzzi, 24 - 10129 Torino (Italy); Canepa, S. [Lab of Reactor Physics and Systems Behaviour LRS, Paul Scherrer Inst., 5232 Villigen (Switzerland)
2012-07-01
The determination of the subcriticality level is an important issue in accelerator-driven system technology. The area method, originally introduced by N. G. Sjoestrand, is a classical technique to interpret flux measurement for pulsed experiments in order to reconstruct the reactivity value. In recent times other methods have also been developed, to account for spatial and spectral effects, which were not included in the area method, since it is based on the point kinetic model. The artificial neural network approach can be an efficient technique to infer reactivities from pulsed experiments. In the present work, some comparisons between the two methods are carried out and discussed. (authors)
Structural Health Monitoring Using Neural Network Based Vibrational System Identification
Sofge, Donald A
2007-01-01
Composite fabrication technologies now provide the means for producing high-strength, low-weight panels, plates, spars and other structural components which use embedded fiber optic sensors and piezoelectric transducers. These materials, often referred to as smart structures, make it possible to sense internal characteristics, such as delaminations or structural degradation. In this effort we use neural network based techniques for modeling and analyzing dynamic structural information for recognizing structural defects. This yields an adaptable system which gives a measure of structural integrity for composite structures.
Communication: Separable potential energy surfaces from multiplicative artificial neural networks
Koch, Werner, E-mail: wkoch@thethirdrock.net; Zhang, Dong H. [State Key Laboratory of Molecular Reaction Dynamics and Center for Theoretical Computational Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 457 Zhongshan Road, Dalian (China)
2014-07-14
We present a potential energy surface fitting scheme based on multiplicative artificial neural networks. It has the sum of products form required for efficient computation of the dynamics of multidimensional quantum systems with the multi configuration time dependent Hartree method. Moreover, it results in analytic potential energy matrix elements when combined with quantum dynamics methods using Gaussian basis functions, eliminating the need for a local harmonic approximation. Scaling behavior with respect to the complexity of the potential as well as the requested accuracy is discussed.
Reconstruction of Flaw Profiles Using Neural Networks and Multi-Frequency Eddy Current System
Chady, T.; Caryk, M. [Technical University of Szczecin, al. Piastow 19, 70-310 Szczecin (Poland)
2005-04-09
The objective of this paper is to identify profiles of flaws in conducting plates. To solve this problem, application of a multi-frequency eddy current system (MFES) and artificial neural networks is proposed. Dynamic feed-forward neural networks with various architectures are investigated. Extended experiments with all neural models are carried out in order to select the most promising configuration. Data utilized for the experiments were obtained from the measurements performed on the Inconel plates with EDM flaws.
Logistic Regression and Artificial Neural Networks for Classification of Ovarian Tumors
Logistic Regression and Artificial Neural Networks for Classification of Ovarian Tumors C. Lu1 , J to generate and evaluate both logistic regression models and artificial neural network (ANN) models to predict, including explorative univariate and multivariate analysis, and the development of the logistic regression
Using Feedforward Neural Networks and Forward Selection of Input Variables for an Ergonomics Data
Kaber, David B.
Using Feedforward Neural Networks and Forward Selection of Input Variables for an Ergonomics Data-gradient algorithm to develop an FNN. This article presents an incremental step in the use of FNNs for ergonomics enhancing the effectiveness of the use of neural networks when observations are missing from ergonomics
Fingerprinting Localization based on Neural Networks and Ultra-wideband signals
Paris-Sud XI, Université de
by processing an IR-UWB indoor propagation measurement campaign. The construction of the neural networks and different sizes of the fingerprinting database. Index Terms--Fingerprinting, Localization, Neural networks]. A good indoor localization system must have a high accuracy, a short training phase, a low cost
Real-time Control of a Tokamak Plasma Using Neural Networks
Bishop, Christopher M.
Real-time Control of a Tokamak Plasma Using Neural Networks Christopher M. Bishop , Paul S. Haynes of neural networks for real-time control of the high temperature plasma in a tokamak fusion experiment. The tokamak is currently the principal experimental device for research into the magnetic confinement approach
A NEURAL NETWORK UNDERLYING INDIVIDUAL DIFFERENCES IN EMOTION AND AGGRESSION IN MALE GOLDEN HAMSTERS
Delville, Yvon
A NEURAL NETWORK UNDERLYING INDIVIDUAL DIFFERENCES IN EMOTION AND AGGRESSION IN MALE GOLDEN, Austin, TX 78712, USA Abstract--In rodents, aggressive behavior can be altered by experimental a common neural network. Male golden hamsters were first screened for offensive aggression. Then
Neural network analysis of strength and ductility of welding alloys for high strength low
Cambridge, University of
Neural network analysis of strength and ductility of welding alloys for high strength low alloy There are considerable demands for the development of weld metals for high strength low alloy steels. To assist in meeting such demands, a neural network was trained and tested on a set of data obtained on weld metals
GAS ANALYSIS SYSTEM COMPOSED OF A SOLID-STATE SENSOR ARRAY AND HYBRID NEURAL NETWORK
Osowski, Stanislaw
1 GAS ANALYSIS SYSTEM COMPOSED OF A SOLID-STATE SENSOR ARRAY AND HYBRID NEURAL NETWORK STRUCTURE of the solid state sensor array used for the gas analysis. The applied neural network is composed of two parts of the gas components. The obtained results have shown that the array of partially selective sensors
LONGITUDINAL MULTIPLE SCLEROSIS LESION SEGMENTATION USING 3D CONVOLUTIONAL NEURAL NETWORKS
Ramanathan, M.
LONGITUDINAL MULTIPLE SCLEROSIS LESION SEGMENTATION USING 3D CONVOLUTIONAL NEURAL NETWORKS Suthirth with dice scores comparable to the inter-rater variability. Index Terms-- Multiple Sclerosis, 3D Convolutional Neural Networks, Deep Learning, Neuroimaging 1. INTRODUCTION Multiple Sclerosis (MS) is a chronic
International Joint Conference on Neural Networks Deductive and Inductive Learning in a
Faisal, Kanaan Abed
of connectionist, adaptive (neural) networks. A backpropagation neural network simulator, which features a logistic function that computes values in the range of -1 to +1, is being used in this work. The ultimate goal A deterministic parse applies rules to a stack and buffer of constituents to generate and perform actions on those
Verleysen, Michel
Prediction of visual perceptions with artificial neural networks in a visual prosthesis based visual prosthesis in order to restore partial vision to the blind. In this paper, an attempt been published in [1]. We propose to use artificial neural networks (ANNs) for predicting the features
An Investigation of Artificial Neural Network Architectures in Artificial Life Implementations
Güngör, Tunga
. They use vision, smell and sound as input to their artificial neural network brains, which utilize HebbianAn Investigation of Artificial Neural Network Architectures in Artificial Life Implementations¨C2D!EF4¦GH¤G¦H¦I!46PRQSUTVXW Abstract. In this paper, an artificially created world is defined
DeepCough: A Deep Convolutional Neural Network in A Wearable Cough Detection System
Odam, Kofi
DeepCough: A Deep Convolutional Neural Network in A Wearable Cough Detection System Justice Amoh that employs a wearable acoustic sensor and a deep convolutional neural network for detecting coughs. We evaluate the performance of our system on 14 healthy volunteers and compare it to that of other cough
Evolving Neural Networks for Hexapod Leg Controllers Gary B. Parker and Zhiyi Li
Parker, Gary B.
Evolving Neural Networks for Hexapod Leg Controllers Gary B. Parker and Zhiyi Li Computer Science -- The incremental evolution of neural networks to control hexapod robot locomotion can be separated into two main that produces leg cycles for a hexapod robot. The robot has 12 servo effectors; two per leg to produce
Non-parametric regression and neural-network inll drilling recovery models for carbonate reservoirs
Valkó, Peter
Non-parametric regression and neural-network in®ll drilling recovery models for carbonate This work introduces non-parametric regression and neural network models for forecasting the in®ll drilling variables. This situation mandates proper selection of independent variables for the in®ll drilling recovery
Neural Networks and Expert Systems to solve the problems of large amounts of Experimental Data at JET
Mellit, A; Hadj-Arab, A; Guessoum, A
2004-01-01
An adaptive artificial neural network model for sizing stand-alone photovoltaic systems: Application for isolated sites in Algeria
Tutorial: Neural networks and their potential application in nuclear power plants
Uhrig, R.E. (Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering Oak Ridge National Lab., TN (United States))
1989-01-01
A neural network is a data processing system consisting of a number of simple, highly interconnected processing elements in an architecture inspired by the structure of the cerebral cortex portion of the brain. Hence, neural networks are often capable of doing things which humans or animals do well but which conventional computers often do poorly. Neural networks have emerged in the past few years as an area of unusual opportunity for research, development and application to a variety of real world problems. Indeed, neural networks exhibit characteristics and capabilities not provided by any other technology. Examples include reading Japanese Kanji characters and human handwriting, reading a typewritten manuscript aloud, compensating for alignment errors in robots, interpreting very noise'' signals (e.g. electroencephalograms), modeling complex systems that cannot be modelled mathematically, and predicting whether proposed loans will be good or fail. This paper presents a brief tutorial on neural networks and describes research on the potential applications to nuclear power plants.
On the deduction of galaxy abundances with evolutionary neural networks
Michael Taylor; Angeles I. Diaz
2007-09-19
A growing number of indicators are now being used with some confidence to measure the metallicity(Z) of photoionisation regions in planetary nebulae, galactic HII regions(GHIIRs), extra-galactic HII regions(EGHIIRs) and HII galaxies(HIIGs). However, a universal indicator valid also at high metallicities has yet to be found. Here, we report on a new artificial intelligence-based approach to determine metallicity indicators that shows promise for the provision of improved empirical fits. The method hinges on the application of an evolutionary neural network to observational emission line data. The network's DNA, encoded in its architecture, weights and neuron transfer functions, is evolved using a genetic algorithm. Furthermore, selection, operating on a set of 10 distinct neuron transfer functions, means that the empirical relation encoded in the network solution architecture is in functional rather than numerical form. Thus the network solutions provide an equation for the metallicity in terms of line ratios without a priori assumptions. Tapping into the mathematical power offered by this approach, we applied the network to detailed observations of both nebula and auroral emission lines in the optical for a sample of 96 HII-type regions and we were able to obtain an empirical relation between Z and S23 with a dispersion of only 0.16 dex. We show how the method can be used to identify new diagnostics as well as the nonlinear relationship supposed to exist between the metallicity Z, ionisation parameter U and effective (or equivalent) temperature T*.
Neural Networks in press URL ftp://iserv.iki.kfki.hu/pub/papers/new/szepes.cc.ps.Z
Szepesvari, Csaba
Neural Networks in press URL ftp://iserv.iki.kfki.hu/pub/papers/new/szepes.cc.ps.Z WWW http. Keywords: Neural network control, compensating perturbations, stability, feedback control, feedforward March 1997 #12; method. 1 Introduction A vast amount of work has dealt with neural networks
Kurchan, Jorge
1 Neural Networks During the 80's and early 90's, there was a great interest in the statistical mechanics community in `Neural Network' models. This interest was centered around, on one hand the Hopfield increasingly clear that the days of Neural Networks as a pure Statistical Mechanical exercise were over
Choi, Seungjin
Cascade Neural Networks for Multichannel Blind Deconvolution Seungjin CHOI \\Lambda 1 and Andrzej deconvolution/equalization, neural networks, unsupervised learning algorithms. Appeared in Electronics Letters. In this letter, we present an efficient online extraction method using cascade neural networks which can extract
Surface daytime net radiation estimation using artificial neural networks
DOE Public Access Gateway for Energy & Science Beta (PAGES Beta)
Jiang, Bo; Zhang, Yi; Liang, Shunlin; Zhang, Xiaotong; Xiao, Zhiqiang
2014-11-11
Net all-wave surface radiation (Rn) is one of the most important fundamental parameters in various applications. However, conventional Rn measurements are difficult to collect because of the high cost and ongoing maintenance of recording instruments. Therefore, various empirical Rn estimation models have been developed. This study presents the results of two artificial neural network (ANN) models (general regression neural networks (GRNN) and Neuroet) to estimate Rn globally from multi-source data, including remotely sensed products, surface measurements, and meteorological reanalysis products. Rn estimates provided by the two ANNs were tested against in-situ radiation measurements obtained from 251 global sites between 1991–2010more »both in global mode (all data were used to fit the models) and in conditional mode (the data were divided into four subsets and the models were fitted separately). Based on the results obtained from extensive experiments, it has been proved that the two ANNs were superior to linear-based empirical models in both global and conditional modes and that the GRNN performed better and was more stable than Neuroet. The GRNN estimates had a determination coefficient (R2) of 0.92, a root mean square error (RMSE) of 34.27 W·m–2 , and a bias of –0.61 W·m–2 in global mode based on the validation dataset. In conclusion, ANN methods are a potentially powerful tool for global Rn estimation.« less
BIAS-FREE SHEAR ESTIMATION USING ARTIFICIAL NEURAL NETWORKS
Gruen, D.; Seitz, S.; Koppenhoefer, J.; Riffeser, A., E-mail: dgruen@usm.uni-muenchen.d [University Observatory Munich, Scheinerstrasse 1, 81679 Muenchen (Germany)
2010-09-01
Bias due to imperfect shear calibration is the biggest obstacle when constraints on cosmological parameters are to be extracted from large area weak lensing surveys such as Pan-STARRS-3{pi}, DES, or future satellite missions like EUCLID. We demonstrate that bias present in existing shear measurement pipelines (e.g., KSB) can be almost entirely removed by means of neural networks. In this way, bias correction can depend on the properties of the individual galaxy instead of being a single global value. We present a procedure to train neural networks for shear estimation and apply this to subsets of simulated GREAT08 RealNoise data. We also show that circularization of the point-spread function (PSF) before measuring the shear reduces the scatter related to the PSF anisotropy correction and thus leads to improved measurements, particularly on low and medium signal-to-noise data. Our results are competitive with the best performers in the GREAT08 competition, especially for the medium and higher signal-to-noise sets. Expressed in terms of the quality parameter defined by GREAT08, we achieve a Q{approx} 40, 140, and 1300 without and 50, 200, and 1300 with circularization for low, medium, and high signal-to-noise data sets, respectively.
Monitoring nuclear reactor systems using neural networks and fuzzy logic
Ikonomopoulos, A.; Tsoukalas, L.H.; Uhrig, R.E. [Tennessee Univ., Knoxville, TN (United States); Mullens, J.A. [Tennessee Univ., Knoxville, TN (United States)]|[Oak Ridge National Lab., TN (United States)
1991-12-01
A new approach is presented that demonstrates the potential of trained artificial neural networks (ANNs) as generators of membership functions for the purpose of monitoring nuclear reactor systems. ANN`s provide a complex-to-simple mapping of reactor parameters in a process analogous to that of measurement. Through such ``virtual measurements`` the value of parameters with operational significance, e.g., control-valve-disk-position, valve-line-up or performance can be determined. In the methodology presented the output of a virtual measuring device is a set of membership functions which independently represent different states of the system. Utilizing a fuzzy logic representation offers the advantage of describing the state of the system in a condensed form, developed through linguistic descriptions and convenient for application in monitoring, diagnostics and generally control algorithms. The developed methodology is applied to the problem of measuring the disk position of the secondary flow control valve of an experimental reactor using data obtained during a start-up. The enhanced noise tolerance of the methodology is clearly demonstrated as well as a method for selecting the actual output. The results suggest that it is possible to construct virtual measuring devices through artificial neural networks mapping dynamic time series to a set of membership functions and thus enhance the capability of monitoring systems. 8 refs., 11 figs., 1 tab.
Yao, Kun
2015-01-01
We demonstrate a convolutional neural network trained to reproduce the Kohn-Sham kinetic energy of hydrocarbons from electron density. The output of the network is used as a non-local correction to the conventional local and semi-local kinetic functionals. We show that this approximation qualitatively reproduces Kohn-Sham potential energy surfaces when used with conventional exchange correlation functionals. Numerical noise inherited from the non-linearity of the neural network is identified as the major challenge for the model. Finally we examine the features in the density learned by the neural network to anticipate the prospects of generalizing these models.
Allman, John M.
IEEE TRANSACTIONS ON NEURAL NETWORKS. VOL. 2, NO. 2, MARCH 1991 231 A Real-Time Neural System for Color Constancy Andrew Moore, Student Member, IEEE, John Allman, and Rodney M. Goodman, Member, IEEE and Neural Systems Program at the California Institute of Technology, Pasadena, CA 91125. IEEE Log Number
An evaluation of neural networks for identification of system parameters in reactor noise signals
Miller, L.F.
1991-12-31
Several backpropagation neural networks for identifying fundamental mode eigenvalues were evaluated. The networks were trained and tested on analytical data and on results from other numerical methods. They were then used to predict first mode break frequencies for noise data from several sources. These predictions were, in turn, compared with analytical values and with results from alternative methods. Comparisons of results for some data sets suggest that the accuracy of predictions from neural networks are essentially equivalent to results from conventional methods while other evaluations indicate that either method may be superior. Experience gained from these numerical experiments provide insight for improving the performance of neural networks relative to other methods for identifying parameters associated with experimental data. Neural networks may also be used in support of conventional algorithms by providing starting points for nonlinear minimization algorithms.
An evaluation of neural networks for identification of system parameters in reactor noise signals
Miller, L.F.
1991-01-01
Several backpropagation neural networks for identifying fundamental mode eigenvalues were evaluated. The networks were trained and tested on analytical data and on results from other numerical methods. They were then used to predict first mode break frequencies for noise data from several sources. These predictions were, in turn, compared with analytical values and with results from alternative methods. Comparisons of results for some data sets suggest that the accuracy of predictions from neural networks are essentially equivalent to results from conventional methods while other evaluations indicate that either method may be superior. Experience gained from these numerical experiments provide insight for improving the performance of neural networks relative to other methods for identifying parameters associated with experimental data. Neural networks may also be used in support of conventional algorithms by providing starting points for nonlinear minimization algorithms.
NOVELTY DETECTION USING AUTO-ASSOCIATIVE NEURAL NETWORK
H. SOHN; K. WORDEN; C. FARRAR
2001-05-01
The primary objective of novelty detection is to examine if a system significantly deviates from the initial baseline condition of the system. In reality, the system is often subject to changing environmental and operation conditions affecting its dynamic characteristics. Such variations include changes in loading, boundary conditions, temperature, and humidity. Most damage diagnosis techniques, however, generally neglect the effects of these changing ambient conditions. Here, a novelty detection technique is developed explicitly taking into account these natural variations of the system in order to minimize false positive indications of true system changes. Auto-associative neural networks are employed to discriminate system changes of interest such as structural deterioration and damage from the natural variations of the system.
Two-Photon Exchange Effect Studied with Neural Networks
Krzysztof M. Graczyk
2011-08-30
An approach to the extraction of the two-photon exchange (TPE) correction from elastic $ep$ scattering data is presented. The cross section, polarization transfer (PT), and charge asymmetry data are considered. It is assumed that the TPE correction to the PT data is negligible. The form factors and TPE correcting term are given by one multidimensional function approximated by the feed forward neural network (NN). To find a model-independent approximation the Bayesian framework for the NNs is adapted. A large number of different parametrizations is considered. The most optimal model is indicated by the Bayesian algorithm. The obtained fit of the TPE correction behaves linearly in epsilon but it has a nontrivial Q2 dependence. A strong dependence of the TPE fit on the choice of parametrization is observed.
Rotation-invariant convolutional neural networks for galaxy morphology prediction
Dieleman, Sander; Dambre, Joni
2015-01-01
Measuring the morphological parameters of galaxies is a key requirement for studying their formation and evolution. Surveys such as the Sloan Digital Sky Survey (SDSS) have resulted in the availability of very large collections of images, which have permitted population-wide analyses of galaxy morphology. Morphological analysis has traditionally been carried out mostly via visual inspection by trained experts, which is time-consuming and does not scale to large ($\\gtrsim10^4$) numbers of images. Although attempts have been made to build automated classification systems, these have not been able to achieve the desired level of accuracy. The Galaxy Zoo project successfully applied a crowdsourcing strategy, inviting online users to classify images by answering a series of questions. Unfortunately, even this approach does not scale well enough to keep up with the increasing availability of galaxy images. We present a deep neural network model for galaxy morphology classification which exploits translational and r...
Regular graphs maximize the variability of random neural networks
Gilles Wainrib; Mathieu Galtier
2014-12-17
In this work we study the dynamics of systems composed of numerous interacting elements interconnected through a random weighted directed graph, such as models of random neural networks. We develop an original theoretical approach based on a combination of a classical mean-field theory originally developed in the context of dynamical spin-glass models, and the heterogeneous mean-field theory developed to study epidemic propagation on graphs. Our main result is that, surprisingly, increasing the variance of the in-degree distribution does not result in a more variable dynamical behavior, but on the contrary that the most variable behaviors are obtained in the regular graph setting. We further study how the dynamical complexity of the attractors is influenced by the statistical properties of the in-degree distribution.
Two-photon exchange effect studied with neural networks
Graczyk, Krzysztof M. [Institute of Theoretical Physics, University of Wroclaw, pl. M. Borna 9, PL-50-204 Wroclaw (Poland)
2011-09-15
An approach to the extraction of the two-photon exchange (TPE) correction from elastic ep scattering data is presented. The cross-section, polarization transfer (PT), and charge asymmetry data are considered. It is assumed that the TPE correction to the PT data is negligible. The form factors and TPE correcting term are given by one multidimensional function approximated by the feedforward neural network (NN). To find a model-independent approximation, the Bayesian framework for the NNs is adapted. A large number of different parametrizations is considered. The most optimal model is indicated by the Bayesian algorithm. The obtained fit of the TPE correction behaves linearly in {epsilon} but it has a nontrivial Q{sup 2} dependence. A strong dependence of the TPE fit on the choice of parametrization is observed.
APPLICATION OF NEURAL NETWORK ALGORITHMS FOR BPM LINEARIZATION
Musson, John C.; Seaton, Chad; Spata, Mike F.; Yan, Jianxun
2012-11-01
Stripline BPM sensors contain inherent non-linearities, as a result of field distortions from the pickup elements. Many methods have been devised to facilitate corrections, often employing polynomial fitting. The cost of computation makes real-time correction difficult, particulalry when integer math is utilized. The application of neural-network technology, particularly the multi-layer perceptron algorithm, is proposed as an efficient alternative for electrode linearization. A process of supervised learning is initially used to determine the weighting coefficients, which are subsequently applied to the incoming electrode data. A non-linear layer, known as an ?activation layer,? is responsible for the removal of saturation effects. Implementation of a perceptron in an FPGA-based software-defined radio (SDR) is presented, along with performance comparisons. In addition, efficient calculation of the sigmoidal activation function via the CORDIC algorithm is presented.
Nuclear power plant status diagnostics using artificial neural networks
Bartlett, E.B. [Iowa State Univ. of Science and Technology, Ames, IA (United States). Dept. of Mechanical Engineering] [Iowa State Univ. of Science and Technology, Ames, IA (United States). Dept. of Mechanical Engineering; Uhrig, R.E. [Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering] [Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering
1991-12-31
In this work, the nuclear power plant operating status recognition issue is investigated using artificial neural networks (ANNs). The objective is to train an ANN to classify nuclear power plant accident conditions and to assess the potential of future work in the area of plant diagnostics with ANNS. To this end, an ANN was trained to recognize normal operating conditions as well as potentially unsafe conditions based on nuclear power plant training simulator generated accident scenarios. These scenarios include; hot and cold leg loss of coolant, control rod ejection, loss of offsite power, main steam line break, main feedwater line break and steam generator tube leak accidents. Findings show that ANNs can be used to diagnose and classify nuclear power plant conditions with good results.
Nuclear power plant status diagnostics using artificial neural networks
Bartlett, E.B. (Iowa State Univ. of Science and Technology, Ames, IA (United States). Dept. of Mechanical Engineering); Uhrig, R.E. (Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering)
1991-01-01
In this work, the nuclear power plant operating status recognition issue is investigated using artificial neural networks (ANNs). The objective is to train an ANN to classify nuclear power plant accident conditions and to assess the potential of future work in the area of plant diagnostics with ANNS. To this end, an ANN was trained to recognize normal operating conditions as well as potentially unsafe conditions based on nuclear power plant training simulator generated accident scenarios. These scenarios include; hot and cold leg loss of coolant, control rod ejection, loss of offsite power, main steam line break, main feedwater line break and steam generator tube leak accidents. Findings show that ANNs can be used to diagnose and classify nuclear power plant conditions with good results.
Extended Kalman Filter Based Neural Networks Controller For Hot Strip Rolling mill
Moussaoui, A. K. [Electrical Engineering Laboratory of Guelma (LGEG), BP.401, University of Guelma, 24000 (Algeria); Abbassi, H. A.; Bouazza, S. [Universite Badji Mokhtar BP 12--23000-Annaba Algerie (Algeria)
2008-06-12
The present paper deals with the application of an Extended Kalman filter based adaptive Neural-Network control scheme to improve the performance of a hot strip rolling mill. The suggested Neural Network model was implemented using Bayesian Evidence based training algorithm. The control input was estimated iteratively by an on-line extended Kalman filter updating scheme basing on the inversion of the learned neural networks model. The performance of the controller is evaluated using an accurate model estimated from real rolling mill input/output data, and the usefulness of the suggested method is proved.
Feed Forward Neural Network and Optimal Control Problem with Control and State Constraints
Kmet', Tibor [Department of Informatics, Constantine the Philosopher University, Tr. A. Hlinku 1, 949 74 Nitra (Slovakia); Kmet'ova, Maria [Department of Mathematics, Constantine the Philosopher University, Tr. A. Hlinku 1, 949 74 Nitra (Slovakia)
2009-09-09
A feed forward neural network based optimal control synthesis is presented for solving optimal control problems with control and state constraints. The paper extends adaptive critic neural network architecture proposed by [5] to the optimal control problems with control and state constraints. The optimal control problem is transcribed into a nonlinear programming problem which is implemented with adaptive critic neural network. The proposed simulation method is illustrated by the optimal control problem of nitrogen transformation cycle model. Results show that adaptive critic based systematic approach holds promise for obtaining the optimal control with control and state constraints.
Surface daytime net radiation estimation using artificial neural networks
Jiang, Bo; Zhang, Yi; Liang, Shunlin; Zhang, Xiaotong; Xiao, Zhiqiang
2014-11-11
Net all-wave surface radiation (R_{n}) is one of the most important fundamental parameters in various applications. However, conventional R_{n} measurements are difficult to collect because of the high cost and ongoing maintenance of recording instruments. Therefore, various empirical R_{n} estimation models have been developed. This study presents the results of two artificial neural network (ANN) models (general regression neural networks (GRNN) and Neuroet) to estimate R_{n} globally from multi-source data, including remotely sensed products, surface measurements, and meteorological reanalysis products. R_{n} estimates provided by the two ANNs were tested against in-situ radiation measurements obtained from 251 global sites between 1991–2010 both in global mode (all data were used to fit the models) and in conditional mode (the data were divided into four subsets and the models were fitted separately). Based on the results obtained from extensive experiments, it has been proved that the two ANNs were superior to linear-based empirical models in both global and conditional modes and that the GRNN performed better and was more stable than Neuroet. The GRNN estimates had a determination coefficient (R^{2}) of 0.92, a root mean square error (RMSE) of 34.27 W·m^{–2} , and a bias of –0.61 W·m^{–2} in global mode based on the validation dataset. In conclusion, ANN methods are a potentially powerful tool for global R_{n} estimation.
Noise reduction in state space using the Focused Gamma Neural Network
Slatton, Clint
Noise reduction in state space using the Focused Gamma Neural Network Jose C. Principe, Jyh In this paper we utilize the gamma neural model to improve the signal to noise ratio (SNR) of broadband signals corrupted by white noise. The projection of a noisy signal onto the signal subspace can not remove the noise
Alternatives to Energy FunctionBased Analysis of Recurrent Neural Networks
Hassoun, Mohamad H.
1 Alternatives to Energy FunctionBased Analysis of Recurrent Neural Networks Mohamad H. Hassoun Engineering Wayne State University Detroit, MI 48202 Introduction Recently, there has been renewed interest. This renewed interest is a result of the wealth and variety of applications of recurrent neural nets
Neural Networks, Radial Basis Functions, and Complexity Mark A. Kon 1
Kon, Mark
, at the pace hoped for. The vision of the 1960's, when the current field of artificial intelligence became for the nonexpert to the theory of artificial neural networks as embodied in current versions of feedforward about work in the area of natural neural nets. For a general presentation of the theory of artificial
Neural Networks, Radial Basis Functions, and Complexity Mark A. Kon1
Kon, Mark
, at the pace hoped for. The vision of the 1960's, when the current field of artificial intelligence became for the non-expert to the theory of artificial neural networks as embodied in current versions of feedforward about work in the area of natural neural nets. For a general presentation of the theory of artificial
Neural network predictions with error bars \\Lambda William D. Penny and Stephen J. Roberts
Roberts, Stephen
Neural network predictions with error bars \\Lambda William D. Penny and Stephen J. Roberts Neural, Technology and Medicine, London SW7 2BT., U.K. w.penny@ic.ac.uk, s.j.roberts@ic.ac.uk February 21, 1997
Are Arti cial Neural Networks White Boxes? Eyal Kolman and Michael Margaliot y
Margaliot, Michael
networks, knowledge-based networks, rule ex- traction, rule generation, rule re#12;nement. #3; This work compute membership function values, whereas nodes in the second layer perform T-norm operations. Jang a similar equivalence to develop a transformation of a feedforward neural network with Logistic activation
Analysis and forecast of the capesize bulk carriers shipping market using Artificial Neural Networks
Voudris, Athanasios V
2006-01-01
Investing in the bulk carrier market constitutes a rather risky investment due to the volatility of the bulk carrier freight rates. In this study it is attempted to uncover the benefits of using Artificial Neural Networks ...
Singh, Harkirat
2004-09-30
This work is aimed towards the development of an artificially intelligent search algorithm used in conjunction with an Auto Associative Neural Network (AANN) to help locate and reconstruct faulty sensor inputs in control ...
Satisfiability of logic programming based on radial basis function neural networks
Hamadneh, Nawaf; Sathasivam, Saratha; Tilahun, Surafel Luleseged; Choon, Ong Hong [School of Mathematical Sciences, Universiti Sains Malaysia, 11800 USM, Penang (Malaysia)
2014-07-10
In this paper, we propose a new technique to test the Satisfiability of propositional logic programming and quantified Boolean formula problem in radial basis function neural networks. For this purpose, we built radial basis function neural networks to represent the proportional logic which has exactly three variables in each clause. We used the Prey-predator algorithm to calculate the output weights of the neural networks, while the K-means clustering algorithm is used to determine the hidden parameters (the centers and the widths). Mean of the sum squared error function is used to measure the activity of the two algorithms. We applied the developed technique with the recurrent radial basis function neural networks to represent the quantified Boolean formulas. The new technique can be applied to solve many applications such as electronic circuits and NP-complete problems.
Mining customer credit by using neural network model with logistic regression approach
Kao, Ling-Jing
2001-01-01
. The objective of this research was to investigate the methodologies to mine customer credit history for the bank industry. Particularly, combination of logistic regression model and neural network technique are proposed to determine if the predictive capability...
Comprehensive functional testing and dynamic compensation techniques for Cellular Neural Networks
Grimaila, Michael Russell
1995-01-01
Cellular Neural Networks (CNN's) are analog, non-linear, dynamic systems which are especially well suited for solving problems in the areas of image processing and pattern recognition. State of the art implementations of ...
Data driven process monitoring based on neural networks and classification trees
Zhou, Yifeng
2005-11-01
of this dissertation is to develop process monitoring schemes that can be applied to complex process systems. Neural networks have been a popular tool for modeling and pattern classification for monitoring of process systems. However, due to the prohibitive...
Optoelectronic implementations of Pulse-Coupled Neural Networks : challenges and limitations
Wise, Raydiance (Raydiance Raychele)
2007-01-01
This thesis examines Pulse Coupled Neural Networks (PCNNs) and their applications, and the feasibility of a compact, rugged, cost-efficient optoelectronic implementation. Simulation results are presented. Proposed optical ...
Long, Xiao
2012-10-19
networks, the element with the highest potential to initiate a fatigue crack is identified, and the time elapsed up to the crack initiation is predicted and compared with its true synthetic answer. Following similar procedures, a competitive neural array...
EEGbased communication via dynamic neural network models William D. Penny and Stephen J. Roberts
Roberts, Stephen
EEGbased communication via dynamic neural network models William D. Penny and Stephen J. Roberts fw.penny, s.j.robertsg@ic.ac.uk Department of Electrical and Electronic Engineering, Imperial College
EEGbased communication via dynamic neural network models William D. Penny and Stephen J. Roberts
Penny, Will
EEGÂbased communication via dynamic neural network models William D. Penny and Stephen J. Roberts fw.penny, s.j.robertsg@ic.ac.uk Department of Electrical and Electronic Engineering, Imperial College
Identification of reactor vessel failures using spatiotemporal neural networks
Roh, C.H.; Chang, H.S.; Kim, H.G.; Chang, S.H. [Korea Advanced Inst. of Science and Technology, Taejon (Korea, Republic of). Dept. of Nuclear Engineering
1996-12-01
Identification of vessel failures provides operators and technical support center personnel with important information to manage severe accidents in a nuclear power plant. It may be very difficult, however, for operators to identify a reactor vessel failure simply by watching temporal trends of some parameters because they have not experienced severe accidents. Therefore, the authors propose a methodology on the identification of pressurized water reactor (PWR) vessel failure for severe accident management using spatiotemporal neural network (STN). STN can deal directly with the spatial and temporal aspects of input signals and can well identify a time-varying problem. Target patterns of seven parameter signals were generated for training the network from the modular accident in nuclear power plants. They integrated MAAP code with STN in on-line system to mimic real accident situation in nuclear power plants. Using new pattern of signals that had never been used for training, the identification capability of STN was tested in a real-time manner. At the tests, STN developed in this study demonstrated acceptable performance in identifying the occurrence of a vessel failure. It is found that STN techniques can be extended to the identification of other key events such as onset of core uncovery, coremelt initiation, containment failure, etc.
Electrokinetic confinement of axonal growth for dynamically configurable neural networks
Honegger, Thibault
Axons in the developing nervous system are directed via guidance cues, whose expression varies both spatially and temporally, to create functional neural circuits. Existing methods to create patterns of neural connectivity ...
Chen, X.; Liu, X.; Gales, M. J. F.; Woodland, P. C.
2015-04-22
, recurrent neural network, GPU, noise contrastive estimation, speech recognition 1. INTRODUCTION Statistical language models (LMs) are crucial components in many speech and language processing systems designed for tasks such as speech recognition, spoken... as follows. In section 2, recur- rent neural network LMs are reviewed. Noise contrastive estimation is presented in section 3. The detailed implement of NCE training is presented in section 4. Experiment results on a large vocabulary conversational telephone...
Nuclear mass systematics by complementing the Finite Range Droplet Model with neural networks
S. Athanassopoulos; E. Mavrommatis; K. A. Gernoth; J. W. Clark
2005-11-30
A neural-network model is developed to reproduce the differences between experimental nuclear mass-excess values and the theoretical values given by the Finite Range Droplet Model. The results point to the existence of subtle regularities of nuclear structure not yet contained in the best microscopic/phenomenological models of atomic masses. Combining the FRDM and the neural-network model, we create a hybrid model with improved predictive performance on nuclear-mass systematics and related quantities.
Use of neural networks to identify transient operating conditions in nuclear power plants
Uhrig, R.E.; Guo, Z.
1989-01-01
A technique using neural networks as a means of diagnosing specific abnormal conditions or problems in nuclear power plants is investigated and found to be feasible. The technique is based on the fact that each physical state of the plant can be represented by a unique pattern of instrument readings, which can be related to the condition of the plant. Neural networks are used to relate this pattern to the fault or problem. 3 refs., 2 figs., 4 tabs.
Neural network based short-term load forecasting using weather compensation
Chow, T.W.S.; Leung, C.T. [City Univ. of Hong Kong, Kowloon (Hong Kong). Dept. of Electronic Engineering] [City Univ. of Hong Kong, Kowloon (Hong Kong). Dept. of Electronic Engineering
1996-11-01
This paper presents a novel technique for electric load forecasting based on neural weather compensation. The proposed method is a nonlinear generalization of Box and Jenkins approach for nonstationary time-series prediction. A weather compensation neural network is implemented for one-day ahead electric load forecasting. The weather compensation neural network can accurately predict the change of actual electric load consumption from the previous day. The results, based on Hong Kong Island historical load demand, indicate that this methodology is capable of providing a more accurate load forecast with a 0.9% reduction in forecast error.
Multi-parameter estimating photometric redshifts with artificial neural networks
Lili Li; Yanxia Zhang; Yongheng Zhao; Dawei Yang
2007-04-17
We calculate photometric redshifts from the Sloan Digital Sky Survey Data Release 2 Galaxy Sample using artificial neural networks (ANNs). Different input patterns based on various parameters (e.g. magnitude, color index, flux information) are explored and their performances for redshift prediction are compared. For ANN technique, any parameter may be easily incorporated as input, but our results indicate that using dereddening magnitude produces photometric redshift accuracies often better than the Petrosian magnitude or model magnitude. Similarly, the model magnitude is also superior to Petrosian magnitude. In addition, ANNs also show better performance when the more effective parameters increase in the training set. Finally, the method is tested on a sample of 79, 346 galaxies from the SDSS DR2. When using 19 parameters based on the dereddening magnitude, the rms error in redshift estimation is sigma(z)=0.020184. The ANN is highly competitive tool when compared with traditional template-fitting methods where a large and representative training set is available.
Evidence for single top quark production using Bayesian neural networks
Kau, Daekwang; /Florida State U.
2007-08-01
We present results of a search for single top quark production in p{bar p} collisions using a dataset of approximately 1 fb{sup -1} collected with the D0 detector. This analysis considers the muon+jets and electron+jets final states and makes use of Bayesian neural networks to separate the expected signals from backgrounds. The observed excess is associated with a p-value of 0.081%, assuming the background-only hypothesis, which corresponds to an excess over background of 3.2 standard deviations for a Gaussian density. The p-value computed using the SM signal cross section of 2.9 pb is 1.6%, corresponding to an expected significance of 2.2 standard deviations. Assuming the observed excess is due to single top production, we measure a single top quark production cross section of {sigma}(p{bar p} {yields} tb + X, tqb + X) = 4.4 {+-} 1.5 pb.
Neural network recognition of nuclear power plant transients
Bartlett, E.B.; Danofsky, R.; Adams, J.; AlJundi, T.; Basu, A.; Dhanwada, C.; Kerr, J.; Kim, K.; Lanc, T.
1993-02-23
The objective of this report is to describe results obtained during the first year of funding that will lead to the development of an artificial neural network (ANN) fault - diagnostic system for the real - time classification of operational transients at nuclear power plants. The ultimate goal of this three-year project is to design, build, and test a prototype diagnostic adviser for use in the control room or technical support center at Duane Arnold Energy Center (DAEC); such a prototype could be integrated into the plant process computer or safety - parameter display system. The adviser could then warn and inform plant operators and engineers of plant component failures in a timely manner. This report describes the work accomplished in the first of three scheduled years for the project. Included herein is a summary of the first year's results as, well as individual descriptions of each of the major topics undertaken by the researchers. Also included are reprints of the articles written under this funding as well as those that were published during the funded period.
Modeling of transport phenomena in tokamak plasmas with neural networks
Meneghini, O.; Luna, C. J.; Smith, S. P.; Lao, L. L.
2014-06-15
A new transport model that uses neural networks (NNs) to yield electron and ion heat flux profiles has been developed. Given a set of local dimensionless plasma parameters similar to the ones that the highest fidelity models use, the NN model is able to efficiently and accurately predict the ion and electron heat transport profiles. As a benchmark, a NN was built, trained, and tested on data from the 2012 and 2013 DIII-D experimental campaigns. It is found that NN can capture the experimental behavior over the majority of the plasma radius and across a broad range of plasma regimes. Although each radial location is calculated independently from the others, the heat flux profiles are smooth, suggesting that the solution found by the NN is a smooth function of the local input parameters. This result supports the evidence of a well-defined, non-stochastic relationship between the input parameters and the experimentally measured transport fluxes. The numerical efficiency of this method, requiring only a few CPU-?s per data point, makes it ideal for scenario development simulations and real-time plasma control.
Zuo Guangqing; Ma Jitang; Bo, B. [Luleaa Univ. of Technology (Sweden). Div. of Process Metallurgy
1996-12-31
To predict the silicon content of hot metal at No. 2 blast furnace, SSAB, Luleaa Works, a three-layer Back-Propagation network model has been established. The network consists of twenty-eight inputs, six middle nodes and one output and uses a generalized delta rule for training. Different network structures and different training strategies have been tested. A well-functioning network with dynamic updating has been designed. The off-line test and the on-line application results showed that more than 80% of the predictions can match the actual silicon content in hot metal in a normal operation, if the allowable prediction error was set to {+-}0.05% Si, while the actual fluctuation of the silicon content was larger than {+-}0.10% Si.
q-state Potts-glass neural network based on pseudoinverse rule
Xiong Daxing; Zhao Hong [Department of Physics and Institute of Theoretical Physics and Astrophysics, Xiamen University, Xiamen 361005 (China)
2010-08-15
We study the q-state Potts-glass neural network with the pseudoinverse (PI) rule. Its performance is investigated and compared with that of the counterpart network with the Hebbian rule instead. We find that there exists a critical point of q, i.e., q{sub cr}=14, below which the storage capacity and the retrieval quality can be greatly improved by introducing the PI rule. We show that the dynamics of the neural networks constructed with the two learning rules respectively are quite different; but however, regardless of the learning rules, in the q-state Potts-glass neural networks with q{>=}3 there is a common novel dynamical phase in which the spurious memories are completely suppressed. This property has never been noticed in the symmetric feedback neural networks. Free from the spurious memories implies that the multistate Potts-glass neural networks would not be trapped in the metastable states, which is a favorable property for their applications.
A Hybrid Neural Network and Simulated Annealing Approach to the Unit Commitment Problem
Liang, Huizhi "Elly"
A Hybrid Neural Network and Simulated Annealing Approach to the Unit Commitment Problem R. Nayak1 network and the simulated annealing method to solve unit commitment, a mixed integer combinatorial variables corresponding to the state of each unit at each time interval. The simulated annealing method
Reinforcement Learning, Neural Networks and PI Control Applied to a Heating Coil
Kretchmar, R. Matthew
Reinforcement Learning, Neural Networks and PI Control Applied to a Heating Coil Charles W,along@lamar.colostate.edu Abstract An accurate simulation of a heating coil is used to compare the performance of a PI controller networks and other artificial intelligence techniques to the control of heating and air
EVOLVING NEURAL NETWORK CONTROLLERS TO PRODUCE LEG CYCLES FOR GAIT GENERATION
Parker, Gary B.
@conncoll.edu ABSTRACT The generation of gaits for hexapod locomotion controllers can be divided into two main parts the structure of an artificial neural network (NN) that produces leg cycles in a hexapod robot. The movement network, robot, hexapod, control INTRODUCTION Because movement and exploration of the environment
Peterson, Carsten
May 1990 LU TP 908 Using Neural Networks to Identify Jets Leif L¨onnblad 1 , Carsten Peterson 2 the ancestor of a hadron jet is presented. The idea is to find an efficient mapping between certain observedpropagated through the network. With this method we are able to separate gluon from quark jets originating from Monte
Neurocomputing 70 (2006) 603606 Stability analysis of an unsupervised neural network
Pilyugin, Sergei S.
2006-01-01
and neurons forming complicated dynamics. 2. Equilibrium and global asymptotic stability analysis of neuro-synapticNeurocomputing 70 (2006) 603606 Letters Stability analysis of an unsupervised neural network neuron of a N-neuron network are STM: _xj ¼ Àajxj þ Cj XN i¼1 Dijf ðxiÞ þ Bj XN i¼1 mijyi, (1) LTM 1
Ohga, Yukiharu; Seki, Hiroshi (Hitachi, Ltd. Energy Research Lab., Ibarakiken (Japan))
1993-02-01
The combination of a neural network and knowledge processing have been used to identify abnormal events that cause a reactor to scram in a nuclear power plant. The neural network recognizes the abnormal event from the change pattern of analog data for state variables, and this result is confirmed from digital data using a knowledge base of plant status when each event occurs. The event identification method is tested using test data based on simulated results of a transient analysis program for boiling water reactors. It is confirmed that a neural network can identify an event in which it has been trained even when the plant conditions, such as fuel burnup, differ from those used in the training and when the analog data contain white noise. The network does not mistakenly identify the nontrained event as a trained one. The method is feasible for event identification, and knowledge processing improves the reliability of the identification.
Convergence and Rate Analysis of Neural Networks for Sparse Approximation
Rozell, Christopher J.
, Abstract--We present an analysis of the Locally Competitive Algorithm (LCA), a Hopfield-style neural theories of neural coding and applications in signal processing. However, the LCA lacks analysis of its to the specifics of the LCA architecture. We show that the LCA has desirable convergence properties
Sikora, R.; Chady, T.; Baniukiewicz, P.; Caryk, M.; Piekarczyk, B. [West Pomeranian University of Technology in Szczecin, Department of Electrical Engineering, 70-313 Szczecin (Poland)
2010-02-22
Nondestructive testing and evaluation are under continuous development. Currently researches are concentrated on three main topics: advancement of existing methods, introduction of novel methods and development of artificial intelligent systems for automatic defect recognition (ADR). Automatic defect classification algorithm comprises of two main tasks: creating a defect database and preparing a defect classifier. Here, the database was built using defect features that describe all geometrical and texture properties of the defect. Almost twenty carefully selected features calculated for flaws extracted from real radiograms were used. The radiograms were obtained from shipbuilding industry and they were verified by qualified operator. Two weld defect's classifiers based on artificial neural networks were proposed and compared. First model consisted of one neural network model, where each output neuron corresponded to different defect group. The second model contained five neural networks. Each neural network had one neuron on output and was responsible for detection of defects from one group. In order to evaluate the effectiveness of the neural networks classifiers, the mean square errors were calculated for test radiograms and compared.
Flow Regime Identification of Co-Current Downward Two-Phase Flow With Neural Network Approach
Hiroshi Goda; Seungjin Kim; Ye Mi; Finch, Joshua P.; Mamoru Ishii [Purdue University, West Lafayette, IN 47907 (United States); Jennifer Uhle [U.S. Nuclear Regulatory Commission, Washington, DC 20555-0001 (United States)
2002-07-01
Flow regime identification for an adiabatic vertical co-current downward air-water two-phase flow in the 25.4 mm ID and the 50.8 mm ID round tubes was performed by employing an impedance void meter coupled with the neural network classification approach. This approach minimizes the subjective judgment in determining the flow regimes. The signals obtained by an impedance void meter were applied to train the self-organizing neural network to categorize these impedance signals into a certain number of groups. The characteristic parameters set into the neural network classification included the mean, standard deviation and skewness of impedance signals in the present experiment. The classification categories adopted in the present investigation were four widely accepted flow regimes, viz. bubbly, slug, churn-turbulent, and annular flows. These four flow regimes were recognized based upon the conventional flow visualization approach by a high-speed motion analyzer. The resulting flow regime maps classified by the neural network were compared with the results obtained through the flow visualization method, and consequently the efficiency of the neural network classification for flow regime identification was demonstrated. (authors)
Uhrig, R.E. (Tennessee Univ., Knoxville, TN (United States) Oak Ridge National Lab., TN (United States))
1990-01-01
Application of neural networks to monitoring and decision making in the operation of nuclear power plants is being investigated under a US Department of Energy sponsored program at the University of Tennessee. Projects include the feasibility of using neural networks for the following tasks: (1) diagnosing specific abnormal conditions or problems in nuclear power plants, (2) detection of the change of mode of operation of the plant, (3) validating signals coming from detectors, (4) review of noise'' data from TVA's Sequoyah Nuclear Power Plant, and (5) examination of the NRC's database of Letter Event Reports'' for correlation of sequences of events in the reported incidents. Each of these projects and its status are described briefly in this paper. This broad based program has as its objective the definition of the state-of-the-art in using neural networks to enhance the performance of commercial nuclear power plants.
Use of neural networks in the operation of nuclear power plants
Uhrig, R.E. (Tennessee Univ., Knoxville, TN (USA) Oak Ridge National Lab., TN (USA))
1990-01-01
Application of neural networks to the operation of nuclear power plants is being investigated under a US Department of Energy sponsored program at the University of Tennessee. Projects include the feasibility of using neural networks for the following tasks: (a) diagnosing specific abnormal conditions, (b) detection of the change of mode of operation, (c) signal validation, (d) monitoring of check valves, (e) modeling of the plant thermodynamics, (f) emulation of core reload calculations, (g) analysis of temporal sequences in NRC's licensee event report,'' (h) monitoring of plant parameters, and (i) analysis of plant vibrations. Each of these projects and its status are described briefly in this article. the objective of each of these projects is to enhance the safety and performance of nuclear plants through the use of neural networks. 6 refs.
Bornholdt, Stefan
Neural network interpretation of LWD data (ODP Leg 170) con¢rms complete sediment subduction-While-Drill- ing (LWD) geophysical data in terms of lithology using an arti¢cial neural network. In combination of the underthrust section. LWD data are measured during the drilling process at the frontal part of the drill string
Signoroni, Alberto
Bacterial Colony Counting by Convolutional Neural Networks Alessandro Ferrari1,2, Stefano Lombardi1 Convolutional Neural Networks (CNN) for counting the number of colonies contained in confluent agglomerates of bacterial cultures on Petri dishes to be screened daily. Computer vision tech- niques are optimal candidates
Draghici, Sorin
A neural network based artificial vision system for licence plate recognition Sorin Draghici, Dept based artificial vision system able to analyse the image of a car given by a camera, locate plate, real-world application #12;A neural network based artificial vision system for licence plate
Zhang, Liqing
enhancement, and biomedical signal processing. Several neural-networks and statistical signal-processing or independent component analysis has attracted considerable attention in the signal- processing and neural-network society, since it not only in- troduces a novel paradigm for signal processing, but also has rapidly
Liu, Y. A.
Predictive Modeling of Large-Scale Commercial Water Desalination Plants: Data-Based Neural Network for developing predictive models for large-scale commercial water desalination plants by (1) a data (MSF) and reverse osmosis (RO) desalination plants in the world. Our resulting neural network
Oliveira, Aurélio R. L.
Neural Networks Give a Warm Start to Linear Optimization Problems MARTA I. VELAZCO , AURELIO R Carlos, SP BRAZIL aurelio@icmc.sc.usp.br Abstract - Hopfield neural networks and interior point methods are used in an integrated way to solve linear optimization problems. The neural network unveils a warm
Cambridge, University of
Neural Networks and Genetic Algorithms in Materials Science and Engineering, 2006 January 11, Shibpur, Howrah, India Neural Networks in Materials Science: The Importance of Uncertainty H. K. D. H rela- tionships and structure within vast arrays of illunderstood data. The neural network method
Development of Ensemble Neural Network Convection Parameterizations for Climate Models
Fox-Rabinovitz, M. S.; Krasnopolsky, V. M.
2012-05-02
The novel neural network (NN) approach has been formulated and used for development of a NN ensemble stochastic convection parametrization for climate models. This fast parametrization is built based on data from Cloud Resolving Model (CRM) simulations initialized with and forced by TOGA-COARE data. The SAM (System for Atmospheric Modeling), developed by D. Randall, M. Khairoutdinov, and their collaborators, has been used for CRM simulations. The observational data are also used for validation of model simulations. The SAM-simulated data have been averaged and projected onto the GCM space of atmospheric states to implicitly define a stochastic convection parametrization. This parametrization is emulated using an ensemble of NNs. An ensemble of NNs with different NN parameters has been trained and tested. The inherent uncertainty of the stochastic convection parametrization derived in such a way is estimated. Due to these inherent uncertainties, NN ensemble is used to constitute a stochastic NN convection parametrization. The developed NN convection parametrization have been validated in a diagnostic CAM (CAM-NN) run vs. the control CAM run. Actually, CAM inputs have been used, at every time step of the control/original CAM integration, for parallel calculations of the NN convection parametrization (CAM-NN) to produce its outputs as a diagnostic byproduct. Total precipitation (P) and cloudiness (CLD) time series, diurnal cycles, and P and CLD distributions for the large Tropical Pacific Ocean for the parallel CAM-NN and CAM runs show similarity and consistency with the NCEP reanalysis. The P and CLD distributions for the tropical area for the parallel runs have been analyzed first for the TOGA-COARE boreal winter season (November 1992 through February 1993) and then for the winter seasons of the follow-up parallel decadal simulations. The obtained results are encouraging and practically meaningful. They show the validity of the NN approach. This constitutes an important practical conclusion of the study: the obtained results on NN ensembles as a stochastic physics parametrization show a realistic possibility of development of NN convection parametrization for climate (and NWP) models based on learning cloud physics from CRM/SAM simulated data.
Computing single step operators of logic programming in radial basis function neural networks
Hamadneh, Nawaf; Sathasivam, Saratha; Choon, Ong Hong [School of Mathematical Sciences, Universiti Sains Malaysia, 11800 USM, Penang (Malaysia)
2014-07-10
Logic programming is the process that leads from an original formulation of a computing problem to executable programs. A normal logic program consists of a finite set of clauses. A valuation I of logic programming is a mapping from ground atoms to false or true. The single step operator of any logic programming is defined as a function (T{sub p}:I?I). Logic programming is well-suited to building the artificial intelligence systems. In this study, we established a new technique to compute the single step operators of logic programming in the radial basis function neural networks. To do that, we proposed a new technique to generate the training data sets of single step operators. The training data sets are used to build the neural networks. We used the recurrent radial basis function neural networks to get to the steady state (the fixed point of the operators). To improve the performance of the neural networks, we used the particle swarm optimization algorithm to train the networks.
A Fast Learning Strategy Using Pattern Selection for Feedforward Neural Networks
Paris-Sud XI, Université de
. The algorithm builds the training corpus starting from a small randomly chosen initial dataset and new patterns of the hidden units [8]. A fairly new approach considered in the last decade is the reduction of the size-class and inter-class variations. Hence, neural network scheme based approaches are time costly solutions
Zhang, Xuesong; Liang, Faming; Yu, Beibei; Zong, Ziliang
2011-11-09
Estimating uncertainty of hydrologic forecasting is valuable to water resources and other relevant decision making processes. Recently, Bayesian Neural Networks (BNNs) have been proved powerful tools for quantifying uncertainty of streamflow forecasting. In this study, we propose a Markov Chain Monte Carlo (MCMC) framework to incorporate the uncertainties associated with input, model structure, and parameter into BNNs. This framework allows the structure of the neural networks to change by removing or adding connections between neurons and enables scaling of input data by using rainfall multipliers. The results show that the new BNNs outperform the BNNs that only consider uncertainties associated with parameter and model structure. Critical evaluation of posterior distribution of neural network weights, number of effective connections, rainfall multipliers, and hyper-parameters show that the assumptions held in our BNNs are not well supported. Further understanding of characteristics of different uncertainty sources and including output error into the MCMC framework are expected to enhance the application of neural networks for uncertainty analysis of hydrologic forecasting.
Artificial Neural Networks In Electric Power Industry Technical Report of the ISIS Group
Antsaklis, Panos
Artificial Neural Networks In Electric Power Industry Technical Report of the ISIS Group at the University of Notre Dame ISIS-94-007 April, 1994 Rafael E. Bourguet and Panos J. Antsaklis Department," Technical Report of the ISIS (Interdisciplinary Studies of Intelligent Systems) Group, No. ISIS-94-007, Univ
Annan, Carl Ashie
2014-12-18
(fd) friction in a physical valve model. Each simulation generates controller output OP(t) and process variable PV(t) time series data. A feed-forward neural network (the predictor) is trained to model the relationship between a given OP and PV pattern...
A Neural Network Based Adaptive Sliding Mode Controller: Application to a Power System Stabilizer
Al-Duwaish, Hussain N.
A Neural Network Based Adaptive Sliding Mode Controller: Application to a Power System Stabilizer University of Petroleum & Minerals, Dhahran 31261, Saudi Arabia * Corresponding Author- e-mail: hduwaish gains when the operating point changes. The proposed method has been applied to a power system
An Application of Context-Learning in a Goal-Seeking Neural Network Thomas E. Portegys,
Portegys, Thomas E.
An Application of Context-Learning in a Goal-Seeking Neural Network Thomas E. Portegys, Illinois State University Normal, Illinois 61790 USA portegys@ilstu.edu ABSTRACT An important function of many], and Hasselmo and McClellands' model of the hippocampus' role in memory formation [6]. In the robotics field
Evolving Neural Network Weights for Time-Series Prediction of General Aviation Flight Data
Hu, Wen-Chen
and predictive maintenance systems, reducing accident rates and saving lives. Keywords: Time-Series Prediction and lucrative industry, it has the highest accident rates within civil aviation [21]. For many years between 0.08% for altitude to 2% for roll. Cross validation of the best neural networks indicate
Reservoir characterization using seismic attributes, well data, and artificial neural networks
Toinet, Sylvain
2001-01-01
. The solution to this problem was proposed by Minsky and Papert (1969). ' adding a hidden layer of units improves the representation power of a neural network (e. g. increses its number of degrees of freedom). It has been shown by Hartman, Keeler and Kowalski...
An Evolutionary Approach for Achieving Scalability with General Regression Neural Networks
Garrett, Aaron
as the optimal feature set for a general regression neural network. We compare its performance against a standard approaches are Gerry Dozier is the director of the Applied Computational Intelligence (ACI) Laboratory in the Department of Computer Science and Software Engineering at Auburn University, as well as members of the ACI
NN-OPT: Neural Network for Option Pricing Using Multinomial Tree
Magdon-Ismail, Malik
on historical data. Finally, we illustrate the power of such a framework by developing a real time trading7]. Other related topics can be found in [2, 8]. Option trading by directly predicting prices and then building trading sys- tems based on the predictions have been considered in the neural network liter- ature
Personalized Spell Checking using Neural Networks Tyler Garaas, Mei Xiao, and Marc Pomplun
Pomplun, Marc
Personalized Spell Checking using Neural Networks Tyler Garaas, Mei Xiao, and Marc Pomplun Visual., Boston, MA 02125-3393, USA tgaraas@cs.umb.edu, meixiao@cs.umb.edu, marc@cs.umb.edu Abstract. Spell;2 Tyler Garaas, Mei Xiao, and Marc Pomplun must be performed to transform one word into another, combined
Neural NetworkBased Modeling and Optimization for Effective Vehicle Emission Testing and
Huang, Yinlun
Introduction Automotive emission of hydrocarbons (HC), carbon monoxide (CO), and nitrogen oxides (NOx) has beenNeural NetworkBased Modeling and Optimization for Effective Vehicle Emission Testing and Engine emission testing and engine calibration are the key to achieving emission standards with satisfactory fuel
NEURAL PCA NETWORK FOR LUNG OUTLINE RECONSTRUCTION IN VQ SCAN IMAGES
Serpen, Gursel
NEURAL PCA NETWORK FOR LUNG OUTLINE RECONSTRUCTION IN VQ SCAN IMAGES G. Serpen1 , Ph. D., R. Iyer1 system takes the digitized ventilation-perfusion scan images of lungs as input, identify a template according to the size and shape of the lungs and thereby approximate and reconstruct the outline of the lung
Enhance Computational Efficiency of Neural Network Predictive Control Using PSO with
Li, Yangmin
Enhance Computational Efficiency of Neural Network Predictive Control Using PSO with Controllable velocity (PSO-CREV), to re- place of GDA in NNPC. Therefore for one cycle of control, PSO-CREV needs less iterations than GDA, and less population size than conven- tional PSO. Hence the computational cost of NNPC
Neural Network Training Using Stochastic PSO Xin Chen and Yangmin Li
Li, Yangmin
Neural Network Training Using Stochastic PSO Xin Chen and Yangmin Li Department is huge, when PSO algorithms are applied for NN training, the dimension of search space is so large that PSOs always converge prematurely. In this paper an improved stochastic PSO (SPSO) is presented
Anderson, Charles W.
Synthesis of Reinforcement Learning, Neural Networks, and PI Control Applied to a Simulated Heating to the control of heating and air-conditioning systems.4,5,8 In this article, three approaches to improving) controlleris appliedto the simulated heating coil and the controller's proportional and integral gains are set
Protein Fold Class Prediction using Neural Networks with Tailored Early-Stopping
Igel, Christian
Protein Fold Class Prediction using Neural Networks with Tailored Early-Stopping Thomas fold class given the primary sequence of a protein. Different feature spaces for primary sequences the fold class of proteins [8], [9], [10], [11], [12], [13], [14], [15], [16], [17]. A fold class contains
Rucci, Michele
Manufacturing cell formation with production data using neural networks R. Sudhakara Pandian, S Exceptional elements a b s t r a c t Batch type production strategies need adoption of cellular manufacturing (CM) in order to improve oper- ational effectiveness by reducing manufacturing lead time and costs
Narasimhan, Giri
Neural Network Classifiers and Gene Selection Methods for Microarray Data on Human Lung with lung cancer gene expression data sets available from the CAMDA website. Many different classification several new methods for classifying gene expression data from lung cancer patients. Our approach uses
Oldenburg, Carl von Ossietzky Universität
be applied to predict the profit, market movements, and price level based on the market's historical datasetClassification with Artificial Neural Networks and Support Vector Machines: application to oil, and Oil fluorescence ABSTRACT: This paper reports on oil classification with fluorescence spectroscopy
Neural Networks and Machine Learning Solutions of the Exam of 29/1/2007
Almeida, Luis B.
Neural Networks and Machine Learning Solutions of the Exam of 29/1/2007 Notes: · The solutions they solved the problems was right or not. The "carefully justified manner" requested in the exam required as 1 (top) and 2 (bottom), and the unit of the second layer as 3. We shall designate by si and yi
Coutinho, Alvaro L. G. A.
Durability Assessment of an Arch Dam using Inverse Analysis with Neural Networks and High de Brasília diannemv@guarany.cpd.unb.br Abstract: In the present work, an analysis of the Funil dam, a double curvature arch dam placed in the state of Rio de Janeiro, Brazil, is presented. The considered
Iterative prediction of chaotic time series using a recurrent neural network
Essawy, M.A.; Bodruzzaman, M. [Tennessee State Univ., Nashville, TN (United States). Dept. of Electrical and Computer Engineering; Shamsi, A.; Noel, S. [USDOE Morgantown Energy Technology Center, WV (United States)
1996-12-31
Chaotic systems are known for their unpredictability due to their sensitive dependence on initial conditions. When only time series measurements from such systems are available, neural network based models are preferred due to their simplicity, availability, and robustness. However, the type of neutral network used should be capable of modeling the highly non-linear behavior and the multi-attractor nature of such systems. In this paper the authors use a special type of recurrent neural network called the ``Dynamic System Imitator (DSI)``, that has been proven to be capable of modeling very complex dynamic behaviors. The DSI is a fully recurrent neural network that is specially designed to model a wide variety of dynamic systems. The prediction method presented in this paper is based upon predicting one step ahead in the time series, and using that predicted value to iteratively predict the following steps. This method was applied to chaotic time series generated from the logistic, Henon, and the cubic equations, in addition to experimental pressure drop time series measured from a Fluidized Bed Reactor (FBR), which is known to exhibit chaotic behavior. The time behavior and state space attractor of the actual and network synthetic chaotic time series were analyzed and compared. The correlation dimension and the Kolmogorov entropy for both the original and network synthetic data were computed. They were found to resemble each other, confirming the success of the DSI based chaotic system modeling.
Dawson, Michael
Networks as Analytic Tools in an ERP Study of Face Memory Reiko Graham1 and Michael R.W. Dawson2 1 Center-related potential (ERP) correlates of face memory has yet to be confirmed. We investigated the possibility neural networks (ANN's) and ANOVA in classifying ERP's from right temporal areas elicited by recognized
Evaluation of Neural Networks for Data Classification, Recognition, and Navigation
Wood, Stephen L.
and Environmental Systems at Florida Institute of Technology in partial fulfillment of the requirements for a combined network, but improved results are achieved using a "fish only" network with pulse energy gathering presented. #12;iv Table of Contents List of Keywords
Part-of-Speech Tagging with Recurrent Neural Networks
Forcada, Mikel L.
networks (DTRNN) for part-of-speech (PoS) tagging of ambiguous words from the sequential infor- mation stored in the network's state. PoS tagging [10] is a very important intermediate step in many natural language processing applications. A PoS tagger is a program that assigns each word in a text a PoS tag
Cambridge, University of
in austenitic stainless steel welds M. Vasudevan, M. Murugananth*, and A.K. Bhaduri Materials Joining Section the influence of compositional variations on ferrite content for the austenitic stainless steel base content in austenitic stainless steel welds based on the optimized neural network model. Bayesian neural
A Global Model of $?^-$-Decay Half-Lives Using Neural Networks
N. Costiris; E. Mavrommatis; K. A. Gernoth; J. W. Clark
2007-01-31
Statistical modeling of nuclear data using artificial neural networks (ANNs) and, more recently, support vector machines (SVMs), is providing novel approaches to systematics that are complementary to phenomenological and semi-microscopic theories. We present a global model of $\\beta^-$-decay halflives of the class of nuclei that decay 100% by $\\beta^-$ mode in their ground states. A fully-connected multilayered feed forward network has been trained using the Levenberg-Marquardt algorithm, Bayesian regularization, and cross-validation. The halflife estimates generated by the model are discussed and compared with the available experimental data, with previous results obtained with neural networks, and with estimates coming from traditional global nuclear models. Predictions of the new neural-network model are given for nuclei far from stability, with particular attention to those involved in r-process nucleosynthesis. This study demonstrates that in the framework of the $\\beta^-$-decay problem considered here, global models based on ANNs can at least match the predictive performance of the best conventional global models rooted in nuclear theory. Accordingly, such statistical models can provide a valuable tool for further mapping of the nuclidic chart.
Apyan, Aram
In this paper, a search for supersymmetry (SUSY) is presented in events with two opposite-sign isolated leptons in the final state, accompanied by hadronic jets and missing transverse energy. An artificial neural network ...
Neural Network Based Modeling of a Large Steam Turbine-Generator Rotor Body Parameters from On technique to estimate and model rotor- body parameters of a large steam turbine-generator from real time
van Milligen, Boudewijn
1995-01-01
1 Proc. 10th Int. Conf. on Stellarators, Madrid, EUR-CIEMAT 30 (1995) 49 Solving Equilibria with a Neural Network B.Ph. van Milligen, V. Tribaldos, J.A. Jiménez Asociación EURATOM-CIEMAT para Fusión Avda
Sontag, Eduardo
Reprinted from: Artificial Neural Networks for Speech and Vision (Proc. Workshop held at Rutgers, in order to increase representational bias, by imposing artificial conditions such as asking that certain
Moon, Jin Woo; Chang, Jae D.; Kim, Sooyoung
2013-07-18
This study examines the performance and adaptability of Artificial Neural Network (ANN)-based thermal control strategies for diverse thermal properties of building envelope conditions applied to residential buildings. The thermal performance using...
Seismic Facies Classification And Identification By Competitive Neural Networks
Saggaf, Muhammad M.
2000-01-01
We present an approach based on competitive networks for the classification and identification of reservoir facies from seismic data. This approach can be adapted to perform either classification of the seismic facies based ...
Islam, Mohammed J; Sid-Ahmed, Maher A; 10.5121/ijaia.2010.1301
2010-01-01
In this paper we present an efficient computer aided mass classification method in digitized mammograms using Artificial Neural Network (ANN), which performs benign-malignant classification on region of interest (ROI) that contains mass. One of the major mammographic characteristics for mass classification is texture. ANN exploits this important factor to classify the mass into benign or malignant. The statistical textural features used in characterizing the masses are mean, standard deviation, entropy, skewness, kurtosis and uniformity. The main aim of the method is to increase the effectiveness and efficiency of the classification process in an objective manner to reduce the numbers of false-positive of malignancies. Three layers artificial neural network (ANN) with seven features was proposed for classifying the marked regions into benign and malignant and 90.91% sensitivity and 83.87% specificity is achieved that is very much promising compare to the radiologist's sensitivity 75%.
AN ARTIFICIAL NEURAL NETWORK EVALUATION OF TUBERCULOSIS USING GENETIC AND PHYSIOLOGICAL PATIENT DATA
Griffin, William O.; Darsey, Jerry A. [Department of Chemistry of Arkansas at Little Rock, Little Rock, AR (United States); Hanna, Josh [Department of Bioinformatics of Arkansas at Little Rock, Little Rock, AR (United States); Razorilova, Svetlana; Kitaev, Mikhael; Alisherov, Avtandiil [National Center of Tuberculosis, Bishkek (Kyrgyzstan); Tarasenko, Olga [Department of Biology University of Arkansas at Little Rock, Little Rock, AR (United States)
2010-04-12
When doctors see more cases of patients with tell-tale symptoms of a disease, it is hoped that they will be able to recognize an infection administer treatment appropriately, thereby speeding up recovery for sick patients. We hope that our studies can aid in the detection of tuberculosis by using a computer model called an artificial neural network. Our model looks at patients with and without tuberculosis (TB). The data that the neural network examined came from the following: patient' age, gender, place, of birth, blood type, Rhesus (Rh) factor, and genes of the human Leukocyte Antigens (HLA) system (9q34.1) present in the Major Histocompatibility Complex. With availability in genetic data and good research, we hope to give them an advantage in the detection of tuberculosis. We try to mimic the doctor's experience with a computer test, which will learn from patient data the factors that contribute to TB.
Utama, R; Prosper, H B
2015-01-01
Besides their intrinsic nuclear-structure value, nuclear mass models are essential for astrophysical applications, such as r-process nucleosynthesis and neutron-star structure. To overcome the intrinsic limitations of existing "state-of-the-art" mass models, we propose a refinement based on a Bayesian Neural Network (BNN) formalism. A novel BNN approach is implemented with the goal of optimizing mass residuals between theory and experiment. A significant improvement (of about 40%) in the mass predictions of existing models is obtained after BNN refinement. Moreover, these improved results are now accompanied by proper statistical errors. Finally, by constructing a "world average" of these predictions, a mass model is obtained that is used to predict the composition of the outer crust of a neutron star. The power of the Bayesian neural network method has been successfully demonstrated by a systematic improvement in the accuracy of the predictions of nuclear masses. Extension to other nuclear observables is a n...
Li, Jun; Jiang, Bin; Guo, Hua, E-mail: hguo@unm.edu [Department of Chemistry and Chemical Biology, University of New Mexico, Albuquerque, New Mexico 87131 (United States)] [Department of Chemistry and Chemical Biology, University of New Mexico, Albuquerque, New Mexico 87131 (United States)
2013-11-28
A rigorous, general, and simple method to fit global and permutation invariant potential energy surfaces (PESs) using neural networks (NNs) is discussed. This so-called permutation invariant polynomial neural network (PIP-NN) method imposes permutation symmetry by using in its input a set of symmetry functions based on PIPs. For systems with more than three atoms, it is shown that the number of symmetry functions in the input vector needs to be larger than the number of internal coordinates in order to include both the primary and secondary invariant polynomials. This PIP-NN method is successfully demonstrated in three atom-triatomic reactive systems, resulting in full-dimensional global PESs with average errors on the order of meV. These PESs are used in full-dimensional quantum dynamical calculations.
A Neural Network Model for Construction Projects Site Overhead Cost Estimating in Egypt
ElSawy, Ismaail; Razek, Mohammed Abdel
2011-01-01
Estimating of the overhead costs of building construction projects is an important task in the management of these projects. The quality of construction management depends heavily on their accurate cost estimation. Construction costs prediction is a very difficult and sophisticated task especially when using manual calculation methods. This paper uses Artificial Neural Network (ANN) approach to develop a parametric cost-estimating model for site overhead cost in Egypt. Fifty-two actual real-life cases of building projects constructed in Egypt during the seven year period 2002-2009 were used as training materials. The neural network architecture is presented for the estimation of the site overhead costs as a percentage from the total project price.
Orozco-Monteagudo, Maykel; Taboada-Crispi, Alberto; Gutierrez-Hernandez, Liliana
2008-11-06
This paper deals with the controversial topic of the selection of the parameters of a genetic algorithm, in this case hierarchical, used for training of multilayer perceptron neural networks for the binary classification. The parameters to select are the crossover and mutation probabilities of the control and parametric genes and the permanency percent. The results can be considered as a guide for using this kind of algorithm.
Tripathi, M.M.; Upadhyay, K.G.; Singh, S.N.
2008-11-15
For the economic and secure operation of power systems, a precise short-term load forecasting technique is essential. Modern load forecasting techniques - especially artificial neural network methods - are particularly attractive, as they have the ability to handle the non-linear relationships between load, weather temperature, and the factors affecting them directly. A test of two different ANN models on data from Australia's Victoria market is promising. (author)
Neural network predictions for Z' boson within LEP2 data set of Bhabha process
A. N. Buryk; V. V. Skalozub
2009-05-15
The neural network approach is applied to search for the Z'-boson within the LEP2 data set for e+ e- -> e+ e- scattering process. In the course of the analysis, the data set is reduced by 20 percent. The axial-vector and vector couplings of the Z' are estimated at 95 percent CL within a two-parameter fit. The mass is determined to be 0.53-1.05 TeV. Comparisons with other results are given.
Towards neuro-memory-chip: Imprinting multiple memories in cultured neural networks
Baruchi, Itay; Ben-Jacob, Eshel [School of Physics and Astronomy, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel-Aviv University, Tel-Aviv 69978 (Israel)
2007-05-15
We show that using local chemical stimulations it is possible to imprint persisting (days) multiple memories (collective modes of neuron firing) in the activity of cultured neural networks. Microdroplets of inhibitory antagonist are injected at a location selected based on real-time analysis of the recorded activity. The neurons at the stimulated locations turn into a focus for initiating synchronized bursting events (the collective modes) each with its own specific spatiotemporal pattern of neuron firing.
Search for Single Top Quark Production at the D0 Experiment using Bayesian Neural Networks
Andres J. Tanasijczuk
2009-05-30
We present the methodology used to measure the single top quark production cross section in the D0 experiment, and show as an example the results that led to the first evidence of single top quark production in D0 at the Fermilab Tevatron proton-antiproton collider. The selected events are mostly backgrounds, which we separate from the expected signals using three multivariate analysis techniques, one of them being Bayesian neural networks, which we will describe here.
Method of 'optimum observables' and implementation of neural networks in physics investigations
Boos, E. E.; Bunichev, V. E.; Dudko, L. V.; Markina, A. A. [Moscow State University, Institute of Nuclear Physics (Russian Federation)
2008-02-15
A separation of a signal of various physics processes from an overwhelming background is one of the most important problems in contemporary high-energy physics. The application of various multivariate statistical methods, such as the neural-network method, has become one of the popular steps toward optimizing relevant analyses. The choice of optimum variables that would disclose distinctions between a signal and a background is one of the important elements in the application of neural networks. A universal method for determining an optimum set of such kinematical variables is described in the present article. The method is based on an analysis of Feynman diagrams contributing to signal and background processes. This method was successfully implemented in searches for single top-quark production with the D0 detector (Tevatron, Fermilab) in analyzing Run I and Run II data. Brief recommendations concerning an optimum implementation of the neural-network method in physics analysis are given on the basis of experience gained in searches for single top-quark production with the D0 detector.
Zhang, Xuesong; Zhao, Kaiguang
2012-06-01
Bayesian Neural Networks (BNNs) have been shown as useful tools to analyze modeling uncertainty of Neural Networks (NNs). This research focuses on the comparison of two BNNs. The first BNNs (BNN-I) use statistical methods to describe the characteristics of different uncertainty sources (input, parameter, and model structure) and integrate these uncertainties into a Markov Chain Monte Carlo (MCMC) framework to estimate total uncertainty. The second BNNs (BNN-II) lump all uncertainties into a single error term (i.e. the residual between model prediction and measurement). In this study, we propose a simple BNN-II, which use Genetic Algorithms (GA) and Bayesian Model Averaging (BMA) to calibrate Neural Networks with different structures (number of hidden units) and combine the predictions from different NNs to derive predictions and uncertainty analysis. We tested these two BNNs in two watersheds for daily and monthly hydrologic simulation. The BMA based BNNs developed in this study outperforms BNN-I in the two watersheds in terms of both accurate prediction and uncertainty estimation. These results show that, given incomplete understanding of the characteristics associated with each uncertainty source, the simple lumped error approach may yield better prediction and uncertainty estimation.
Closed loop adaptive control of spectrum-producing step using neural networks
Fu, C.Y.
1998-11-24
Characteristics of the plasma in a plasma-based manufacturing process step are monitored directly and in real time by observing the spectrum which it produces. An artificial neural network analyzes the plasma spectrum and generates control signals to control one or more of the process input parameters in response to any deviation of the spectrum beyond a narrow range. In an embodiment, a plasma reaction chamber forms a plasma in response to input parameters such as gas flow, pressure and power. The chamber includes a window through which the electromagnetic spectrum produced by a plasma in the chamber, just above the subject surface, may be viewed. The spectrum is conducted to an optical spectrometer which measures the intensity of the incoming optical spectrum at different wavelengths. The output of optical spectrometer is provided to an analyzer which produces a plurality of error signals, each indicating whether a respective one of the input parameters to the chamber is to be increased or decreased. The microcontroller provides signals to control respective controls, but these lines are intercepted and first added to the error signals, before being provided to the controls for the chamber. The analyzer can include a neural network and an optional spectrum preprocessor to reduce background noise, as well as a comparator which compares the parameter values predicted by the neural network with a set of desired values provided by the microcontroller. 7 figs.
Closed loop adaptive control of spectrum-producing step using neural networks
Fu, Chi Yung (San Francisco, CA)
1998-01-01
Characteristics of the plasma in a plasma-based manufacturing process step are monitored directly and in real time by observing the spectrum which it produces. An artificial neural network analyzes the plasma spectrum and generates control signals to control one or more of the process input parameters in response to any deviation of the spectrum beyond a narrow range. In an embodiment, a plasma reaction chamber forms a plasma in response to input parameters such as gas flow, pressure and power. The chamber includes a window through which the electromagnetic spectrum produced by a plasma in the chamber, just above the subject surface, may be viewed. The spectrum is conducted to an optical spectrometer which measures the intensity of the incoming optical spectrum at different wavelengths. The output of optical spectrometer is provided to an analyzer which produces a plurality of error signals, each indicating whether a respective one of the input parameters to the chamber is to be increased or decreased. The microcontroller provides signals to control respective controls, but these lines are intercepted and first added to the error signals, before being provided to the controls for the chamber. The analyzer can include a neural network and an optional spectrum preprocessor to reduce background noise, as well as a comparator which compares the parameter values predicted by the neural network with a set of desired values provided by the microcontroller.
Time-of-flight discrimination between gamma-rays and neutrons by neural networks
Serkan Akkoyun
2012-08-13
In gamma-ray spectroscopy, a number of neutrons are emitted from the nuclei together with the gamma-rays and these neutrons influence gamma-ray spectra. An obvious method of separating between neutrons and gamma-rays is based on the time-of-flight (tof) technique. This work aims obtaining tof distributions of gamma-rays and neutrons by using feed-forward artificial neural network (ANN). It was shown that, ANN can correctly classify gamma-ray and neutron events. Testing of trained networks on experimental data clearly shows up tof discrimination of gamma-rays and neutrons.
Stochastic mean field formulation of the dynamics of diluted neural networks
D. Angulo-Garcia; A. Torcini
2014-09-26
We consider pulse-coupled Leaky Integrate-and-Fire neural networks with randomly distributed synaptic couplings. This random dilution induces fluctuations in the evolution of the macroscopic variables and deterministic chaos at the microscopic level. Our main aim is to mimic the effect of the dilution as a noise source acting on the dynamics of a globally coupled non-chaotic system. Indeed, the evolution of a diluted neural network can be well approximated as a fully pulse coupled network, where each neuron is driven by a mean synaptic current plus additive noise. These terms represent the average and the fluctuations of the synaptic currents acting on the single neurons in the diluted system. The main microscopic and macroscopic dynamical features can be retrieved with this stochastic approximation. Furthermore, the microscopic stability of the diluted network can be also reproduced, as demonstrated from the almost coincidence of the measured Lyapunov exponents in the deterministic and stochastic cases for an ample range of system sizes. Our results strongly suggest that the fluctuations in the synaptic currents are responsible for the emergence of chaos in this class of pulse coupled networks.
Neural Network Probability Estimation for Broad Coverage Parsing
(Ratnaparkhi, 1999; Collins, 1999; Charniak, 2001) are based on a history-based probability model (Black et al-crafted finite set of features to represent the unbounded parse history (Ratna- parkhi, 1999; Collins, 1999 Sim- ple Synchrony Networks (SSNs) (Lane and Hen- derson, 2001; Henderson, 2000). Because
Broad Absorption Line Quasar catalogues with Supervised Neural Networks
Simone Scaringi; Christopher E. Cottis; Christian Knigge; Michael R. Goad
2008-10-24
We have applied a Learning Vector Quantization (LVQ) algorithm to SDSS DR5 quasar spectra in order to create a large catalogue of broad absorption line quasars (BALQSOs). We first discuss the problems with BALQSO catalogues constructed using the conventional balnicity and/or absorption indices (BI and AI), and then describe the supervised LVQ network we have trained to recognise BALQSOs. The resulting BALQSO catalogue should be substantially more robust and complete than BI- or AI-based ones.
Artificial neural networks for input-output dynamic modeling of nonlinear processes
Sarimveis, Haralambos
1992-01-01
ARTIFICIAL NEURAL NETWORKS FOR INPUT-OUTPUT DYNAMIC MODELING OF NONLINEAR PROCESSES A Th& vis HARALA'&IBOS SARIlcIVEIS Snhnuttect to th&e Otftc& of Cire&11&nt&' Stll&heb of T& x?. Akrr I L'rrrrc& re& tv in pnrtinl frrlfilhrreut of the rc turr...&iveil as to style anel e&a&t& nt hy. 'liM 1 i'll&. hael Nil&a)sou (Chaii of Comuiitte& ) A Teil Watson (I&lenih&a ) Al& ~ii&)e& G. Parloa ( Ileiubei. ) Rayiinai&1 W. F'lumcifelt ( H& a&1 of Defa& it &neat ) A. u ust 10l1a ABSTRACT Artih& ial Neural...
Prediction of U-Mo dispersion nuclear fuels with Al-Si alloy using artificial neural network
Susmikanti, Mike; Sulistyo, Jos
2014-09-30
Dispersion nuclear fuels, consisting of U-Mo particles dispersed in an Al-Si matrix, are being developed as fuel for research reactors. The equilibrium relationship for a mixture component can be expressed in the phase diagram. It is important to analyze whether a mixture component is in equilibrium phase or another phase. The purpose of this research it is needed to built the model of the phase diagram, so the mixture component is in the stable or melting condition. Artificial neural network (ANN) is a modeling tool for processes involving multivariable non-linear relationships. The objective of the present work is to develop code based on artificial neural network models of system equilibrium relationship of U-Mo in Al-Si matrix. This model can be used for prediction of type of resulting mixture, and whether the point is on the equilibrium phase or in another phase region. The equilibrium model data for prediction and modeling generated from experimentally data. The artificial neural network with resilient backpropagation method was chosen to predict the dispersion of nuclear fuels U-Mo in Al-Si matrix. This developed code was built with some function in MATLAB. For simulations using ANN, the Levenberg-Marquardt method was also used for optimization. The artificial neural network is able to predict the equilibrium phase or in the phase region. The develop code based on artificial neural network models was built, for analyze equilibrium relationship of U-Mo in Al-Si matrix.
Bayesian neural networks for classification: how useful is the evidence W.D. Penny1,*, S.J. Roberts
Penny, Will
Bayesian neural networks for classification: how useful is the evidence framework? W.D. Penny1,*, S: 44-171-823- 8125. E-mail address: w.penny@ic.ac.uk; URL: http://www.ee.ic.ac.uk/ research/neural/wpenny.html (W.D. Penny) 1 W.D. Penny is a research fellow funded by the UK Engineering and Physical Sciences
2009-07-13
number not for citation purposes) BMC Neuroscience Open AccessOral presentation Neural networks with small-world topology are optimal for encoding based on spatiotemporal patterns of spikes Petra E Vertes*1 and Tom Duke2 Address: 1Cavendish Laboratory... - chronous groups, that have been shown to emerge spontaneously in networks of spiking neurons with axonal conduction delays and spike-timing-dependent plasticity. Here, we investigate the effect of network topology on the ease and reliability with which...
Isaksson, Marcus; Jalden, Joakim; Murphy, Martin J.
2005-12-15
In this study we address the problem of predicting the position of a moving lung tumor during respiration on the basis of external breathing signals--a technique used for beam gating, tracking, and other dynamic motion management techniques in radiation therapy. We demonstrate the use of neural network filters to correlate tumor position with external surrogate markers while simultaneously predicting the motion ahead in time, for situations in which neither the breathing pattern nor the correlation between moving anatomical elements is constant in time. One pancreatic cancer patient and two lung cancer patients with mid/upper lobe tumors were fluoroscopically imaged to observe tumor motion synchronously with the movement of external chest markers during free breathing. The external marker position was provided as input to a feed-forward neural network that correlated the marker and tumor movement to predict the tumor position up to 800 ms in advance. The predicted tumor position was compared to its observed position to establish the accuracy with which the filter could dynamically track tumor motion under nonstationary conditions. These results were compared to simplified linear versions of the filter. The two lung cancer patients exhibited complex respiratory behavior in which the correlation between surrogate marker and tumor position changed with each cycle of breathing. By automatically and continuously adjusting its parameters to the observations, the neural network achieved better tracking accuracy than the fixed and adaptive linear filters. Variability and instability in human respiration complicate the task of predicting tumor position from surrogate breathing signals. Our results show that adaptive signal-processing filters can provide more accurate tumor position estimates than simpler stationary filters when presented with nonstationary breathing motion.
Mahdi Bazarghan; Ranjan Gupta
2008-04-26
Automated techniques have been developed to automate the process of classification of objects or their analysis. The large datasets provided by upcoming spectroscopic surveys with dedicated telescopes urges scientists to use these automated techniques for analysis of such large datasets which are now available to the community. Sloan Digital Sky Survey (SDSS) is one of such surveys releasing massive datasets. We use Probabilistic Neural Network (PNN) for automatic classification of about 5000 SDSS spectra into 158 spectral type of a reference library ranging from O type to M type stars.
Jesse, Stephen [ORNL; Kalinin, Sergei V [ORNL; Kumar, Amit [ORNL; Ovchinnikov, Oleg S [ORNL; Guo, Senli [ORNL; Griggio, Flavio [ORNL; Trolier-Mckinstry, Susan E [ORNL
2011-01-01
The spatial variability of the polarization dynamics in thin film ferroelectric capacitors was probed by recognition analysis of spatially-resolved spectroscopic data. Switching spectroscopy piezoresponse force microscopy was used to measure local hysteresis loops and map them on a 2D random-bond, random-field Ising model. A neural-network based recognition approach was utilized to analyze the hysteresis loops and their spatial variability. Strong variability is observed in the polarization dynamics around macroscopic cracks due to the modified local elastic and electric boundary conditions, with most pronounced effect on the length scale of ~100 nm away from the crack.
Imaging regenerating bone tissue based on neural networks applied to micro-diffraction measurements
Campi, G.; Pezzotti, G. [Institute of Crystallography, CNR, via Salaria Km 29.300, I-00015, Monterotondo Roma (Italy)] [Institute of Crystallography, CNR, via Salaria Km 29.300, I-00015, Monterotondo Roma (Italy); Fratini, M. [Centro Fermi -Museo Storico della Fisica e Centro Studi e Ricerche 'Enrico Fermi', Roma (Italy)] [Centro Fermi -Museo Storico della Fisica e Centro Studi e Ricerche 'Enrico Fermi', Roma (Italy); Ricci, A. [Deutsches Elektronen-Synchrotron DESY, Notkestraße 85, D-22607 Hamburg (Germany)] [Deutsches Elektronen-Synchrotron DESY, Notkestraße 85, D-22607 Hamburg (Germany); Burghammer, M. [European Synchrotron Radiation Facility, B. P. 220, F-38043 Grenoble Cedex (France)] [European Synchrotron Radiation Facility, B. P. 220, F-38043 Grenoble Cedex (France); Cancedda, R.; Mastrogiacomo, M. [Istituto Nazionale per la Ricerca sul Cancro, and Dipartimento di Medicina Sperimentale dell'Università di Genova and AUO San Martino Istituto Nazionale per la Ricerca sul Cancro, Largo R. Benzi 10, 16132, Genova (Italy)] [Istituto Nazionale per la Ricerca sul Cancro, and Dipartimento di Medicina Sperimentale dell'Università di Genova and AUO San Martino Istituto Nazionale per la Ricerca sul Cancro, Largo R. Benzi 10, 16132, Genova (Italy); Bukreeva, I.; Cedola, A. [Institute for Chemical and Physical Process, CNR, c/o Physics Dep. at Sapienza University, P-le A. Moro 5, 00185, Roma (Italy)] [Institute for Chemical and Physical Process, CNR, c/o Physics Dep. at Sapienza University, P-le A. Moro 5, 00185, Roma (Italy)
2013-12-16
We monitored bone regeneration in a tissue engineering approach. To visualize and understand the structural evolution, the samples have been measured by X-ray micro-diffraction. We find that bone tissue regeneration proceeds through a multi-step mechanism, each step providing a specific diffraction signal. The large amount of data have been classified according to their structure and associated to the process they came from combining Neural Networks algorithms with least square pattern analysis. In this way, we obtain spatial maps of the different components of the tissues visualizing the complex kinetic at the base of the bone regeneration.
Braunsmann, Christoph; Schäffer, Tilman E., E-mail: tilman.schaeffer@uni-tuebingen.de [Institute of Applied Physics and LISA, University of Tübingen, Auf der Morgenstelle 10, 72076 Tübingen (Germany)
2014-05-15
Force curves recorded with the atomic force microscope on structured samples often show an irregular force versus indentation behavior. An analysis of such curves using standard contact models (e.g., the Sneddon model) would generate inaccurate Young's moduli. A critical inspection of the force curve shape is therefore necessary for estimating the reliability of the generated Young's modulus. We used a trained artificial neural network to automatically recognize curves of “good” and of “bad” quality. This is especially useful for improving the analysis of force maps that consist of a large number of force curves.
Implementation Of The Artificial Neural Networks To Control The Springback Of Metal Sheets
Crina, Axinte [University of Bacau, Calea Marasesti 157, 600115 Bacau (Romania)
2007-05-17
Geometrical inaccuracy of sheet metal parts due to the springback phenomenon is the reason for considerable efforts in tools and process development. Prediction of springback is a key issue to design the tools and control the process parameters in order to obtain close tolerances in the formed parts. The objective of this paper is to use simulation procedure coupled with neural networks method to get the best relation between process parameters and tools geometry in order to minimize the shape deviations of the formed parts related to the target geometry.
Quantitative analysis of tin alloy combined with artificial neural network prediction
Oh, Seong Y.; Yueh, Fang-Yu; Singh, Jagdish P.
2010-05-01
Laser-induced breakdown spectroscopy was applied to quantitative analysis of three impurities in Sn alloy. The impurities analysis was based on the internal standard method using the Sn I 333.062-nm line as the reference line to achieve the best reproducible results. Minor-element concentrations (Ag, Cu, Pb) in the alloy were comparatively evaluated by artificial neural networks (ANNs) and calibration curves. ANN was found to effectively predict elemental concentrations with a trend of nonlinear growth due to self-absorption. The limits of detection for Ag, Cu, and Pb in Sn alloy were determined to be 29, 197, and 213 ppm, respectively.
Improving the training and evaluation efficiency of recurrent neural network language models
Chen, X.; Liu, X.; Gales, M. J. F.; Woodland, P. C.
2015-04-22
in Section 5. Experimental results on a conversational telephone speech transcription task are given in Sec- tion 6 and conclusions presented in Section 7. 2. RECURRENT NEURAL NETWORK LMS In contrast to feedforward NNLMs, recurrent NNLMs [1] represent... , NNLMs are often linearly in- terpolated with n-gram LMs to obtain both a good context coverage and strong generalisation [16, 17, 18, 1, 5, 19]. The interpolated LM probability is given by P (wi|hi?11 ) = ?PNG(wi|hi?11 ) + (1 ? ?)PRNN(wi|hi?11 ) (2) ?...
Zeng, Xiaosi
2011-02-22
system. The model development process can be used as a demonstration of how, in detail, to predict travel time under various freeway conditions by using the neural network approach. The modeling results may be integrated directly into major TMCs... outputs. Since freeway segments are typically much shorter, the pairing of input and outputs become less erroneous and the sample size is usually no longer an issue. Furthermore, the first step of the two-stage method can be used as a building block so...
Systematics on ground-state energies of nuclei within the neural networks
Tuncay Bayram; Serkan Akkoyun; S. Okan Kara
2013-01-11
One of the fundamental ground-state properties of nuclei is binding energy. In this study, we have employed artificial neural networks (ANNs) to obtain binding energies based on the data calculated from Hartree-Fock-Bogolibov (HFB) method with the two SLy4 and SKP Skyrme forces. Also, ANNs have been employed to obtain two-neutron and two-proton separation energies of nuclei. Statistical modeling of nuclear data using ANNs has been seen as to be successful in this study. Such a statistical model can be possible tool for searching in systematics of nuclei beyond existing experimental nuclear data.
Neural Network Technology as a Pollution Prevention Tool in the Electric Utility Industry
Johnson, M. L.
1998-01-01
of more than 6 percent a year, a reduction of more than I million tons of CO 2 per year will be required by the year 2000 to return to 1990 CO 2 levels. The LCRA actively has pursued the use of wind and hydroelectric power, and energy conservation... data. The technical review indicated that the use of a neural network system would require some upgrades to input measurement nodes on devices that are identified as critical for achieving boiler optimization. The ability to implement new...
Uhrig, R.E. [Tennessee Univ., Knoxville, TN (United States)]|[Oak Ridge National Lab., TN (United States)
1990-12-31
Application of neural networks to monitoring and decision making in the operation of nuclear power plants is being investigated under a US Department of Energy sponsored program at the University of Tennessee. Projects include the feasibility of using neural networks for the following tasks: (1) diagnosing specific abnormal conditions or problems in nuclear power plants, (2) detection of the change of mode of operation of the plant, (3) validating signals coming from detectors, (4) review of ``noise`` data from TVA`s Sequoyah Nuclear Power Plant, and (5) examination of the NRC`s database of ``Letter Event Reports`` for correlation of sequences of events in the reported incidents. Each of these projects and its status are described briefly in this paper. This broad based program has as its objective the definition of the state-of-the-art in using neural networks to enhance the performance of commercial nuclear power plants.
Bodruzzaman, M.; Essawy, M.A.
1996-03-31
Chaotic systems are known for their unpredictability due to their sensitive dependence on initial conditions. When only time series measurements from such systems are available, neural network based models are preferred due to their simplicity, availability, and robustness. However, the type of neural network used should be capable of modeling the highly non-linear behavior and the multi- attractor nature of such systems. In this paper we use a special type of recurrent neural network called the ``Dynamic System Imitator (DSI)``, that has been proven to be capable of modeling very complex dynamic behaviors. The DSI is a fully recurrent neural network that is specially designed to model a wide variety of dynamic systems. The prediction method presented in this paper is based upon predicting one step ahead in the time series, and using that predicted value to iteratively predict the following steps. This method was applied to chaotic time series generated from the logistic, Henon, and the cubic equations, in addition to experimental pressure drop time series measured from a Fluidized Bed Reactor (FBR), which is known to exhibit chaotic behavior. The time behavior and state space attractor of the actual and network synthetic chaotic time series were analyzed and compared. The correlation dimension and the Kolmogorov entropy for both the original and network synthetic data were computed. They were found to resemble each other, confirming the success of the DSI based chaotic system modeling.
Schmidhuber, Juergen
Deep Learning in Neural Networks: An Overview Technical Report IDSIA-03-14 / arXiv:1404.7828 v2 [cs In recent years, deep artificial neural networks (including recurrent ones) have won numerous con- tests of it from the previous millennium. Shallow and deep learners are distinguished by the depth of their credit
Schmidhuber, Juergen
Draft: Deep Learning in Neural Networks: An Overview Technical Report IDSIA-03-14 / arXiv:1404 In recent years, deep artificial neural networks (including recurrent ones) have won numerous con- tests of it from the previous millennium. Shallow and deep learners are distinguished by the depth of their credit
Schmidhuber, Juergen
Deep Learning in Neural Networks: An Overview Technical Report IDSIA-03-14 / arXiv:1404.7828v1 [cs In recent years, deep neural networks (including recurrent ones) have won numerous contests in pattern the previous millennium. Shallow and deep learners are distinguished by the depth of their credit assignment
Schmidhuber, Juergen
Deep Learning in Neural Networks: An Overview Technical Report IDSIA-03-14 / arXiv:1404.7828 v4 [cs October 2014 Abstract In recent years, deep artificial neural networks (including recurrent ones) have won relevant work, much of it from the previous millennium. Shallow and deep learners are distin- guished
Schmidhuber, Juergen
Deep Learning in Neural Networks: An Overview Technical Report IDSIA-03-14 / arXiv:1404.7828 v3 [cs In recent years, deep artificial neural networks (including recurrent ones) have won numerous con- tests of it from the previous millennium. Shallow and deep learners are distinguished by the depth of their credit
Alippi, Cesare
solution. Index Terms--Air-fuel ratio control, automotive fuel injection, air pollution, neural network. 2, MAY 2003 259 A Neural-Network Based Control Solution to Air-Fuel Ratio Control for Automotive the design of accurate control sys- tems to keep the air-to-fuel ratio at the optimal stoichiometric value AF
A. Caldwell; F. Cossavella; B. Majorovits; D. Palioselitis; O. Volynets
2015-07-21
A pulse-shape discrimination method based on artificial neural networks was applied to pulses simulated for different background, signal and signal-like interactions inside a germanium detector. The simulated pulses were used to investigate variations of efficiencies as a function of used training set. It is verified that neural networks are well-suited to identify background pulses in true-coaxial high-purity germanium detectors. The systematic uncertainty on the signal recognition efficiency derived using signal-like evaluation samples from calibration measurements is estimated to be 5\\%. This uncertainty is due to differences between signal and calibration samples.
Meyer, B.J.; Sellers, J.P.; Thomsen, J.U.
1993-06-08
Apparatus and processes are described for recognizing and identifying materials. Characteristic spectra are obtained for the materials via spectroscopy techniques including nuclear magnetic resonance spectroscopy, infrared absorption analysis, x-ray analysis, mass spectroscopy and gas chromatography. Desired portions of the spectra may be selected and then placed in proper form and format for presentation to a number of input layer neurons in an offline neural network. The network is first trained according to a predetermined training process; it may then be employed to identify particular materials. Such apparatus and processes are particularly useful for recognizing and identifying organic compounds such as complex carbohydrates, whose spectra conventionally require a high level of training and many hours of hard work to identify, and are frequently indistinguishable from one another by human interpretation.
Meyer, Bernd J. (Athens, GA); Sellers, Jeffrey P. (Suwanee, GA); Thomsen, Jan U. (Fredricksberg, DK)
1993-01-01
Apparatus and processes for recognizing and identifying materials. Characteristic spectra are obtained for the materials via spectroscopy techniques including nuclear magnetic resonance spectroscopy, infrared absorption analysis, x-ray analysis, mass spectroscopy and gas chromatography. Desired portions of the spectra may be selected and then placed in proper form and format for presentation to a number of input layer neurons in an offline neural network. The network is first trained according to a predetermined training process; it may then be employed to identify particular materials. Such apparatus and processes are particularly useful for recognizing and identifying organic compounds such as complex carbohydrates, whose spectra conventionally require a high level of training and many hours of hard work to identify, and are frequently indistinguishable from one another by human interpretation.
Sub-daily Statistical Downscaling of Meteorological Variables Using Neural Networks
Kumar, Jitendra [ORNL] [ORNL; Brooks, Bjørn-Gustaf J. [University of Illinois, Urbana-Champaign] [University of Illinois, Urbana-Champaign; Thornton, Peter E [ORNL] [ORNL; Dietze, Michael [University of Illinois, Urbana-Champaign] [University of Illinois, Urbana-Champaign
2012-01-01
A new open source neural network temporal downscaling model is described and tested using CRU-NCEP reanal ysis and CCSM3 climate model output. We downscaled multiple meteorological variables in tandem from monthly to sub-daily time steps while also retaining consistent correlations between variables. We found that our feed forward, error backpropagation approach produced synthetic 6 hourly meteorology with biases no greater than 0.6% across all variables and variance that was accurate within 1% for all variables except atmospheric pressure, wind speed, and precipitation. Correlations between downscaled output and the expected (original) monthly means exceeded 0.99 for all variables, which indicates that this approach would work well for generating atmospheric forcing data consistent with mass and energy conserved GCM output. Our neural network approach performed well for variables that had correlations to other variables of about 0.3 and better and its skill was increased by downscaling multiple correlated variables together. Poor replication of precipitation intensity however required further post-processing in order to obtain the expected probability distribution. The concurrence of precipitation events with expected changes in sub ordinate variables (e.g., less incident shortwave radiation during precipitation events) were nearly as consistent in the downscaled data as in the training data with probabilities that differed by no more than 6%. Our downscaling approach requires training data at the target time step and relies on a weak assumption that climate variability in the extrapolated data is similar to variability in the training data.
R. Utama; J. Piekarewicz; H. B. Prosper
2015-08-25
Besides their intrinsic nuclear-structure value, nuclear mass models are essential for astrophysical applications, such as r-process nucleosynthesis and neutron-star structure. To overcome the intrinsic limitations of existing "state-of-the-art" mass models, we propose a refinement based on a Bayesian Neural Network (BNN) formalism. A novel BNN approach is implemented with the goal of optimizing mass residuals between theory and experiment. A significant improvement (of about 40%) in the mass predictions of existing models is obtained after BNN refinement. Moreover, these improved results are now accompanied by proper statistical errors. Finally, by constructing a "world average" of these predictions, a mass model is obtained that is used to predict the composition of the outer crust of a neutron star. The power of the Bayesian neural network method has been successfully demonstrated by a systematic improvement in the accuracy of the predictions of nuclear masses. Extension to other nuclear observables is a natural next step that is currently under investigation.
Towards a feasible implementation of quantum neural networks using quantum dots
M. V. Altaisky; N. N. Zolnikova; N. E. Kaputkina; V. A. Krylov; Yu. E. Lozovik; N. S. Dattani
2015-03-17
We propose an implementation of quantum neural networks using an array of single-electron quantum dots with dipole-dipole interactions. We demonstrate that this implementation is both feasible and versatile by studying it within the framework of GaAs based quantum dot qubits coupled to a reservoir of acoustic phonons; a system whose decoherence properties have been experimentally and theoretically characterized with meticulous detail, and is considered one of the most accurately understood open quantum systems. Using numerically exact Feynman integral calculations, we have found that the quantum coherence in our neural networks survive for several ns even at liquid nitrogen temperatures (77 K), which is three orders of magnitude higher than current implementations which are based on SQUIDs operating at temperatures in the mK range. Furthermore, the previous quantum dot based proposals required control via manipulating the phonon bath, which is extremely difficult in real experiments. An advantage of our implementation is that it can be easily controlled, since dipole-dipole interaction strengths can be changed via the spacing between the dots and applying external fields.
Dong, X. Y.; De Robertis, M. M., E-mail: xydong@yorku.ca [Physics and Astronomy Department, York University, Toronto, ON M3J 1P3 (Canada)
2013-10-01
This is the second paper of the series Detecting Active Galactic Nuclei Using Multi-filter Imaging Data. In this paper we review shapelets, an image manipulation algorithm, which we employ to adjust the point-spread function (PSF) of galaxy images. This technique is used to ensure the image in each filter has the same and sharpest PSF, which is the preferred condition for detecting AGNs using multi-filter imaging data as we demonstrated in Paper I of this series. We apply shapelets on Canada-France-Hawaii Telescope Legacy Survey Wide Survey ugriz images. Photometric parameters such as effective radii, integrated fluxes within certain radii, and color gradients are measured on the shapelets-reconstructed images. These parameters are used by artificial neural networks (ANNs) which yield: photometric redshift with an rms of 0.026 and a regression R-value of 0.92; galaxy morphological types with an uncertainty less than 2 T types for z ? 0.1; and identification of galaxies as AGNs with 70% confidence, star-forming/starburst (SF/SB) galaxies with 90% confidence, and passive galaxies with 70% confidence for z ? 0.1. The incorporation of ANNs provides a more reliable technique for identifying AGN or SF/SB candidates, which could be very useful for large-scale multi-filter optical surveys that also include a modest set of spectroscopic data sufficient to train neural networks.
Kyungmin Kim; Ian W. Harry; Kari A. Hodge; Young-Min Kim; Chang-Hwan Lee; Hyun Kyu Lee; John J. Oh; Sang Hoon Oh; Edwin J. Son
2015-03-03
We apply a machine learning algorithm, the artificial neural network, to the search for gravitational-wave signals associated with short gamma-ray bursts. The multi-dimensional samples consisting of data corresponding to the statistical and physical quantities from the coherent search pipeline are fed into the artificial neural network to distinguish simulated gravitational-wave signals from background noise artifacts. Our result shows that the data classification efficiency at a fixed false alarm probability is improved by the artificial neural network in comparison to the conventional detection statistic. Therefore, this algorithm increases the distance at which a gravitational-wave signal could be observed in coincidence with a gamma-ray burst. In order to demonstrate the performance, we also evaluate a few seconds of gravitational-wave data segment using the trained networks and obtain the false alarm probability. We suggest that the artificial neural network can be a complementary method to the conventional detection statistic for identifying gravitational-wave signals related to the short gamma-ray bursts.
White, Tony
with infrared distance sensors. A more detailed physics model would be needed, but the neural network training period of time. Each car is mounted with multiple straight-line distance sensors, which provide the input traveled and rely on this for fitness evaluations. Both evolutionary algorithms are well suited
Stillinger, Frank
amino-acid types, A and B, which allows the determination of the glo- bal energy structure for all reported statistical methods [25,26] are now in fact competing effectively with -65% prediction accuracy possible sources of error exist for neural-network prediction of protein structure. First, the experimental
Paris-Sud XI, Université de
Short-term Wind Power Prediction for Offshore Wind Farms - Evaluation of Fuzzy-Neural Network Based of offshore farms and their secure integration to the grid. Modeling the behavior of large wind farms presents the new considerations that have to be made when dealing with large offshore wind farms
Sinha, Sitabhra
for Advanced Scientific Research, Bangalore 560 064, India 3 Machine Intelligence Unit, Indian Statistical Institute of Science, Bangalore 560 012, India 2 Condensed Matter Theory Unit, Jawaharlal Nehru Center Institute, Calcutta 700 035, India Received 14 December 2001; published 28 March 2002 Neural network models
Chow, Mo-Yuen
forms of energy to provide power to other equipment. In the performance of all motor systems, bearings1060 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 47, NO. 5, OCTOBER 2000 Neural-Network-Based Motor Rolling Bearing Fault Diagnosis Bo Li, Student Member, IEEE, Mo-Yuen Chow, Senior Member, IEEE
Barranco, Bernabe Linares
IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 17, NO. 3, MAY 2006 771 On Algorithmic Rate-Coded AER of frames into the spike event-based representation known as the address-event-rep- resentation (AER). In this paper we concentrate on rate-coded AER. The problem is addressed as an algorithmic problem, in which
Slatton, Clint
Target Discrimination in Synthetic Aperture Radar (SAR) using Artificial Neural Networks 1 Target principe@cnel.ufl.edu Abstract: This paper addresses target discrimination in synthetic aperture radar (SAR classification but here the goal is discrimination. We will show that the two applications require different cost
Bolen, Matthew Scott
2005-11-01
applications being filed in Canada and Mexico for LNG import terminals. The EIA (Energy Information Agency) estimates by 2025 that LNG will make up 21% of the total U.S. Natural Gas Supply. This study developed a neural network approach to forecast LNG imports...
Maclin, Rich
for Protein Folding Richard Maclin Jude W. Shavlik Computer Sciences Dept. University of Wisconsin 1210 W learning Theory refinement Neural networks Finite-state automata Protein folding Chou-Fasman algorithm-Fasman algorithm, a method for predicting how globular proteins fold. Empirical evidence shows
Lerner, Boaz
of feature extraction methods based on statistical pattern recognition or on artificial neural networks(x) and by the criteria they have to optimize. Feature extraction methods can be grouped into four categories [4] based, unsupervized methods are the only way to perform feature extraction. In other cases, supervized paradigms
Mohaghegh, Shahab
SPE 31159 Predicting Well Stimulation Results in a Gas Storage Field in the Absence of Reservoir of this methodology to a gas storage field is presented in this paper. The developed neural network can predict for future stimulation treatment in the aforementioned field. INTRODUCTION Gas storage fields have numerous
Mohaghegh, Shahab
in information technologies. Employing computers in the work place, incorporating sophisticated simula- tion Technologies" has a highly dynamic meaning. In recent years, artificial neural networks and fuzzy set theory Technologies." These tools are providing engineers and scientists with the foundation upon which intelligent
Crabbe, Frederick
architecture Frederick L. Crabbe Michael G. Dyer Artificial Intelligence Lab, Computer Science Department The paper presents a neural network architecture (MAXSON) based on secondÂorder connections that can learn architecture that: learns faster than traditional reinforcement learning approaches, generates and applies
Mitnitski, Arnold B.
1 Assessment of Individual Risk of Death Using Self-report Data: an Artificial Neural Network,4 and Kenneth Rockwood, MD1,4 1 Geriatric Medicine Research Unit, Queen Elizabeth II Health Sciences Centre rate over 10 simulations in predicting the probability of individual survival was 79.2 ± 0
Pearlmutter, Barak
Recurrent Neural Networks: A Survey Barak A. Pearlmutter Abstract--- We survey learning algorithms maximally satisfies a complex set of conflicting con straints [2], [3], [4], [5], [6], a system which]. For this reason, if one is interested in solving a particular problem, it would be only prudent to try a va riety
Potter, Don
Polymer property prediction and optimization using neural networks Nilay K. Roy, Walter D. Potter, and David P. Landau Abstract -- Prediction and optimization of polymer properties is a complex and highly non-linear problem with no easy method to predict polymer properties directly and accurately
Fault Diagnosis of Steam Generator Using Signed Directed Graph and Artificial Neural Networks
Aly, Mohamed N. [Nuclear Eng. Department, Fac. of Eng., Alex. Univ., Alex. (Egypt); Hegazy, Hesham N. [Nuclear Power Plants Authority, Cairo (Egypt)
2006-07-01
Diagnosis is a very complex and important task for finding the root cause of faults in nuclear power plants. The objective of this paper is to investigate the feasibility of using the combination of signed directed graph (SDG) and artificial neural networks for fault diagnosis in nuclear power plants especially in U-Tube steam generator. Signed directed graph has been the most widely used form of qualitative based model methods for process fault diagnosis. It is constructed to represent the cause-effect relations among the dynamic process variables. Signed directed graph consists of nodes represent the process variables and branches. The branch represents the qualitative influence of a process variable on the related variable. The main problem in fault diagnosis using the signed directed graph is the unmeasured variables. Therefore, neural networks are used to estimate the values of unmeasured nodes. In this work, different four cases of faults in the steam generator ( SG) have been diagnosed, three of them are single fault and the fourth is multiple fault. The first three faults are by pass valve leakage (Vbp(+)), main feed water valve opening increase (Vfw(+)), main feed water valve opening decrease (Vfw (-)). The fourth fault is a multiple fault where by-pass valve leakage and main feed water valve opening decrease (Vbp(+) and Vfw (-)) in the same time. The used data are collected from a basic principle simulator of pressurized water reactor 925 Mwe. The signed directed graph of the steam generator is constructed to represent the cause-effect relations among SG variables. It consists of 26 nodes represent the SG variables, and 48 branches represent the cause effect relations among this variables. For each fault the values of measured nodes are coming from sensors and the values of unmeasured nodes are coming from the trained neural networks. These values of the nodes are compared by normal values to get the sign of the nodes. The cause-effect graph for each fault is constructed from the steam generator signed directed graph by removing the invalid (normal) nodes and inconsistent branches. Then in the cause-effect graph we search about the node which does not have an input branch. This node is the fault origin node. The result of this work demonstrated that this method can be used in nuclear power plant fault diagnosis. The advantages of this method are, it enables us to diagnose a multi fault, it is not restricted by pre-defined faults, and it is fast method. (authors)
Mahdi Bazarghan
2008-04-17
A Probabilistic Neural Network model has been used for automated classification of ELODIE stellar spectral library consisting of about 2000 spectra into 158 known spectro-luminosity classes. The full spectra with 561 flux bins and a PCA reduced set of 57, 26 and 16 components have been used for the training and test sessions. The results shows a spectral type classification accuracy of 3.2 sub-spectral type and luminosity class accuracy of 2.7 for the full spectra and an accuracy of 3.1 and 2.6 respectively with the PCA set. This technique will be useful for future upcoming large data bases and their rapid classification.
Bazarghan, Mahdi
2008-01-01
A Probabilistic Neural Network model has been used for automated classification of ELODIE stellar spectral library consisting of about 2000 spectra into 158 known spectro-luminosity classes. The full spectra with 561 flux bins and a PCA reduced set of 57, 26 and 16 components have been used for the training and test sessions. The results shows a spectral type classification accuracy of 3.2 sub-spectral type and luminosity class accuracy of 2.7 for the full spectra and an accuracy of 3.1 and 2.6 respectively with the PCA set. This technique will be useful for future upcoming large data bases and their rapid classification.
Optimal system size for complex dynamics in random neural networks near criticality
Wainrib, Gilles, E-mail: wainrib@math.univ-paris13.fr [Laboratoire Analyse Géométrie et Applications, Université Paris XIII, Villetaneuse (France)] [Laboratoire Analyse Géométrie et Applications, Université Paris XIII, Villetaneuse (France); García del Molino, Luis Carlos, E-mail: garciadelmolino@ijm.univ-paris-diderot.fr [Institute Jacques Monod, Université Paris VII, Paris (France)
2013-12-15
In this article, we consider a model of dynamical agents coupled through a random connectivity matrix, as introduced by Sompolinsky et al. [Phys. Rev. Lett. 61(3), 259–262 (1988)] in the context of random neural networks. When system size is infinite, it is known that increasing the disorder parameter induces a phase transition leading to chaotic dynamics. We observe and investigate here a novel phenomenon in the sub-critical regime for finite size systems: the probability of observing complex dynamics is maximal for an intermediate system size when the disorder is close enough to criticality. We give a more general explanation of this type of system size resonance in the framework of extreme values theory for eigenvalues of random matrices.
Complex dynamics of a delayed discrete neural network of two nonidentical neurons
Chen, Yuanlong [Mathematics Department, GuangDong University of Finance, Guangzhou 510521 (China)] [Mathematics Department, GuangDong University of Finance, Guangzhou 510521 (China); Huang, Tingwen [Mathematics Department, Texas A and M University at Qatar, P. O. Box 23874, Doha (Qatar)] [Mathematics Department, Texas A and M University at Qatar, P. O. Box 23874, Doha (Qatar); Huang, Yu, E-mail: stshyu@mail.sysu.edu.cn [Mathematics Department, Sun Yat-Sen University, Guangzhou 510275, People's Republic China (China)] [Mathematics Department, Sun Yat-Sen University, Guangzhou 510275, People's Republic China (China)
2014-03-15
In this paper, we discover that a delayed discrete Hopfield neural network of two nonidentical neurons with self-connections and no self-connections can demonstrate chaotic behaviors. To this end, we first transform the model, by a novel way, into an equivalent system which has some interesting properties. Then, we identify the chaotic invariant set for this system and show that the dynamics of this system within this set is topologically conjugate to the dynamics of the full shift map with two symbols. This confirms chaos in the sense of Devaney. Our main results generalize the relevant results of Huang and Zou [J. Nonlinear Sci. 15, 291–303 (2005)], Kaslik and Balint [J. Nonlinear Sci. 18, 415–432 (2008)] and Chen et al. [Sci. China Math. 56(9), 1869–1878 (2013)]. We also give some numeric simulations to verify our theoretical results.
Reichenbach, Tobias
2015-01-01
Frequency discrimination is a fundamental task of the auditory system. The mammalian inner ear, or cochlea, provides a place code in which different frequencies are detected at different spatial locations. However, a temporal code based on spike timing is also available: action potentials evoked in an auditory-nerve fiber by a low-frequency tone occur at a preferred phase of the stimulus-they exhibit phase locking-and thus provide temporal information about the tone's frequency. In an accompanying psychoacoustic study, and in agreement with previous experiments, we show that humans employ this temporal information for discrimination of low frequencies. How might such temporal information be read out in the brain? Here we demonstrate that recurrent random neural networks in which connections between neurons introduce characteristic time delays, and in which neurons require temporally coinciding inputs for spike initiation, can perform sharp frequency discrimination when stimulated with phase-locked inputs. Alt...
One and two proton separation energies from nuclear mass systematics using neural networks
S. Athanassopoulos; E. Mavrommatis; K. A. Gernoth; J. W. Clark
2005-09-26
We deal with the systematics of one and two proton separation energies as predicted by our latest global model for the masses of nuclides developed with the use of neural networks. Among others, such systematics is useful as input to the astrophysical rp-process and to the one and two proton radioactive studies. Our results are compared with the experimental separation energies referred to in the 2003 Atomic Mass Evaluation and with those evaluated from theoretical models for the masses of nuclides, like the FRDM of Möller et al. and the HFB2 of Pearson et al. We focus in particular on the proton separation energies for nuclides that are involved in the rp-process (29<=Z<=40) but they have not yet been studied experimentally.
Improvement studies on neutron-gamma separation in HPGe detectors by using neural networks
Serkan Akkoyun; Tuncay Bayram; S. Okan Kara
2013-04-11
The neutrons emitted in heavy-ion fusion-evaporation (HIFE) reactions together with the gamma-rays cause unwanted backgrounds in gamma-ray spectra. Especially in the nuclear reactions, where relativistic ion beams (RIBs) are used, these neutrons are serious problem. They have to be rejected in order to obtain clearer gamma-ray peaks. In this study, the radiation energy and three criteria which were previously determined for separation between neutron and gamma-rays in the HPGe detectors have been used in artificial neural network (ANN) for improving of the decomposition power. According to the preliminary results obtained from ANN method, the ratio of neutron rejection has been improved by a factor of 1.27 and the ratio of the lost in gamma-rays has been decreased by a factor of 0.50.
Hybrid expert system - neural network - Fuzzy Logic methodology for transient identification
Ikonomopoulos, A.; Tsoukalas, L.H. (Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering); Uhrig, R.E. (Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering Oak Ridge National Lab., TN (United States))
1991-01-01
A methodology is presented that demonstrates the potential of pretrained artificial neural networks (ANN's) as generators of membership functions for the purpose of transient identification in Nuclear Power Plants (NPP). In order to provide timely concise and task-specific information about the many aspects of the transient and to determine the state of the system based on the interpretation of potentially noisy data, a model-referenced approach is utilized, where pretrained ANNs provide the model. Membership functions -- that condense information about a transient in a form convenient for a rule-based identification system -- are produced through ANN'S. The results demonstrate the extremely good noise-tolerance of ANN's and suggest a new method for transient identification within the framework of Fuzzy Logic.
Hybrid expert system - neural network - Fuzzy Logic methodology for transient identification
Ikonomopoulos, A.; Tsoukalas, L.H. [Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering; Uhrig, R.E. [Tennessee Univ., Knoxville, TN (United States). Dept. of Nuclear Engineering]|[Oak Ridge National Lab., TN (United States)
1991-12-31
A methodology is presented that demonstrates the potential of pretrained artificial neural networks (ANN`s) as generators of membership functions for the purpose of transient identification in Nuclear Power Plants (NPP). In order to provide timely concise and task-specific information about the many aspects of the transient and to determine the state of the system based on the interpretation of potentially noisy data, a model-referenced approach is utilized, where pretrained ANNs provide the model. Membership functions -- that condense information about a transient in a form convenient for a rule-based identification system -- are produced through ANN`S. The results demonstrate the extremely good noise-tolerance of ANN`s and suggest a new method for transient identification within the framework of Fuzzy Logic.
Neural network modelling of thermal stratification in a solar DHW storage
Geczy-Vig, P.; Farkas, I.
2010-05-15
In this study an artificial neural network (ANN) model is introduced for modelling the layer temperatures in a storage tank of a solar thermal system. The model is based on the measured data of a domestic hot water system. The temperatures distribution in the storage tank divided in 8 equal parts in vertical direction were calculated every 5 min using the average 5 min data of solar radiation, ambient temperature, mass flow rate of collector loop, load and the temperature of the layers in previous time steps. The introduced ANN model consists of two parts describing the load periods and the periods between the loads. The identified model gives acceptable results inside the training interval as the average deviation was 0.22 C during the training and 0.24 C during the validation. (author)
AllamehZadeh, Mostafa, E-mail: dibaparima@yahoo.com [International Institute of Earthquake Engineering and Seismology (Iran, Islamic Republic of)
2011-12-15
A Quadratic Neural Networks (QNNs) model has been developed for identifying seismic source classification problem at regional distances using ARMA coefficients determination by Artificial Neural Networks (ANNs). We have devised a supervised neural system to discriminate between earthquakes and chemical explosions with filter coefficients obtained by windowed P-wave phase spectra (15 s). First, we preprocess the recording's signals to cancel out instrumental and attenuation site effects and obtain a compact representation of seismic records. Second, we use a QNNs system to obtain ARMA coefficients for feature extraction in the discrimination problem. The derived coefficients are then applied to the neural system to train and classification. In this study, we explore the possibility of using single station three-component (3C) covariance matrix traces from a priori-known explosion sites (learning) for automatically recognizing subsequent explosions from the same site. The results have shown that this feature extraction gives the best classifier for seismic signals and performs significantly better than other classification methods. The events have been tested, which include 36 chemical explosions at the Semipalatinsk test site in Kazakhstan and 61 earthquakes (mb = 5.0-6.5) recorded by the Iranian National Seismic Network (INSN). The 100% correct decisions were obtained between site explosions and some of non-site events. The above approach to event discrimination is very flexible as we can combine several 3C stations.
Ortiz-Rodriguez, J. M.; Reyes Alfaro, A.; Reyes Haro, A.; Solis Sanches, L. O.; Miranda, R. Castaneda; Cervantes Viramontes, J. M. [Universidad Autonoma de Zacatecas, Unidad Academica de Ingenieria Electrica. Av. Ramon Lopez Velarde 801. Col. Centro Zacatecas, Zac (Mexico); Vega-Carrillo, H. R. [Universidad Autonoma de Zacatecas, Unidad Academica de Ingenieria Electrica. Av. Ramon Lopez Velarde 801. Col. Centro Zacatecas, Zac., Mexico. and Unidad Academica de Estudios Nucleares. C. Cip (Mexico)
2013-07-03
In this work a neutron spectrum unfolding code, based on artificial intelligence technology is presented. The code called ''Neutron Spectrometry and Dosimetry with Artificial Neural Networks and two Bonner spheres'', (NSDann2BS), was designed in a graphical user interface under the LabVIEW programming environment. The main features of this code are to use an embedded artificial neural network architecture optimized with the ''Robust design of artificial neural networks methodology'' and to use two Bonner spheres as the only piece of information. In order to build the code here presented, once the net topology was optimized and properly trained, knowledge stored at synaptic weights was extracted and using a graphical framework build on the LabVIEW programming environment, the NSDann2BS code was designed. This code is friendly, intuitive and easy to use for the end user. The code is freely available upon request to authors. To demonstrate the use of the neural net embedded in the NSDann2BS code, the rate counts of {sup 252}Cf, {sup 241}AmBe and {sup 239}PuBe neutron sources measured with a Bonner spheres system.
Ortiz-Rodriguez, J. M.; Reyes Alfaro, A.; Reyes Haro, A.; Solis Sanches, L. O.; Miranda, R. Castaneda; Cervantes Viramontes, J. M. [Universidad Autonoma de Zacatecas, Unidad Academica de Ingenieria Electrica. Av. Ramon Lopez Velarde 801. Col. Centro Zacatecas, Zac (Mexico); Vega-Carrillo, H. R. [Universidad Autonoma de Zacatecas, Unidad Academica de Ingenieria Electrica. Av. Ramon Lopez Velarde 801. Col. Centro Zacatecas, Zac., Mexico. and Unidad Academica de Estudios Nucleares. C. Cip (Mexico)
2013-07-03
In this work the performance of two neutron spectrum unfolding codes based on iterative procedures and artificial neural networks is evaluated. The first one code based on traditional iterative procedures and called Neutron spectrometry and dosimetry from the Universidad Autonoma de Zacatecas (NSDUAZ) use the SPUNIT iterative algorithm and was designed to unfold neutron spectrum and calculate 15 dosimetric quantities and 7 IAEA survey meters. The main feature of this code is the automated selection of the initial guess spectrum trough a compendium of neutron spectrum compiled by the IAEA. The second one code known as Neutron spectrometry and dosimetry with artificial neural networks (NDSann) is a code designed using neural nets technology. The artificial intelligence approach of neural net does not solve mathematical equations. By using the knowledge stored at synaptic weights on a neural net properly trained, the code is capable to unfold neutron spectrum and to simultaneously calculate 15 dosimetric quantities, needing as entrance data, only the rate counts measured with a Bonner spheres system. Similarities of both NSDUAZ and NSDann codes are: they follow the same easy and intuitive user's philosophy and were designed in a graphical interface under the LabVIEW programming environment. Both codes unfold the neutron spectrum expressed in 60 energy bins, calculate 15 dosimetric quantities and generate a full report in HTML format. Differences of these codes are: NSDUAZ code was designed using classical iterative approaches and needs an initial guess spectrum in order to initiate the iterative procedure. In NSDUAZ, a programming routine was designed to calculate 7 IAEA instrument survey meters using the fluence-dose conversion coefficients. NSDann code use artificial neural networks for solving the ill-conditioned equation system of neutron spectrometry problem through synaptic weights of a properly trained neural network. Contrary to iterative procedures, in neural net approach it is possible to reduce the rate counts used to unfold the neutron spectrum. To evaluate these codes a computer tool called Neutron Spectrometry and dosimetry computer tool was designed. The results obtained with this package are showed. The codes here mentioned are freely available upon request to the authors.
Coolen, ACC "Ton"
governing the evolution of macro- scopic order parameters in the Hopfield [1] neural net- work model near subsequently been applied to other disordered spins systems [4,5], and is understood to be exact at least (i) in the limit where the disorder is removed (i.e., for attractor neural networks away from saturation
Simões, Marcelo Godoy
in three-dimensional images. The results have been successfully verified. The neural network es- timation a battery-powered wireless transmission system for this application. With the elimination of cabling, sig
Hasanhodzic, Jasmina, 1979-
2004-01-01
We revisit the kernel regression based pattern recognition algorithm designed by Lo, Mamaysky, and Wang (2000) to extract nonlinear patterns from the noisy price data, and develop an analogous neural network based one. We ...
Mellit, A.; Benghanme, M.; Arab, A. H.; Guessoum, A.
2004-01-01
The objective of this work is to investigate the Radial Basis Function Neural Networks (RBFN) to identifying and modeling the optimal sizing couples of stand-alone photovoltaic (PV) system using a minimum of input data, These optimal couples allow...
A linear approach for sparse coding by a two-layer neural network
Montalto, Alessandro; Prevete, Roberto
2015-01-01
Many approaches to transform classification problems from non-linear to linear by feature transformation have been recently presented in the literature. These notably include sparse coding methods and deep neural networks. However, many of these approaches require the repeated application of a learning process upon the presentation of unseen data input vectors, or else involve the use of large numbers of parameters and hyper-parameters, which must be chosen through cross-validation, thus increasing running time dramatically. In this paper, we propose and experimentally investigate a new approach for the purpose of overcoming limitations of both kinds. The proposed approach makes use of a linear auto-associative network (called SCNN) with just one hidden layer. The combination of this architecture with a specific error function to be minimized enables one to learn a linear encoder computing a sparse code which turns out to be as similar as possible to the sparse coding that one obtains by re-training the neura...
Singh, Kunwar P., E-mail: kpsingh_52@yahoo.com [Academy of Scientific and Innovative Research, Council of Scientific and Industrial Research, New Delhi (India); Environmental Chemistry Division, CSIR-Indian Institute of Toxicology Research, Post Box 80, Mahatma Gandhi Marg, Lucknow 226 001 (India); Gupta, Shikha; Rai, Premanjali [Academy of Scientific and Innovative Research, Council of Scientific and Industrial Research, New Delhi (India); Environmental Chemistry Division, CSIR-Indian Institute of Toxicology Research, Post Box 80, Mahatma Gandhi Marg, Lucknow 226 001 (India)
2013-10-15
Robust global models capable of discriminating positive and non-positive carcinogens; and predicting carcinogenic potency of chemicals in rodents were developed. The dataset of 834 structurally diverse chemicals extracted from Carcinogenic Potency Database (CPDB) was used which contained 466 positive and 368 non-positive carcinogens. Twelve non-quantum mechanical molecular descriptors were derived. Structural diversity of the chemicals and nonlinearity in the data were evaluated using Tanimoto similarity index and Brock–Dechert–Scheinkman statistics. Probabilistic neural network (PNN) and generalized regression neural network (GRNN) models were constructed for classification and function optimization problems using the carcinogenicity end point in rat. Validation of the models was performed using the internal and external procedures employing a wide series of statistical checks. PNN constructed using five descriptors rendered classification accuracy of 92.09% in complete rat data. The PNN model rendered classification accuracies of 91.77%, 80.70% and 92.08% in mouse, hamster and pesticide data, respectively. The GRNN constructed with nine descriptors yielded correlation coefficient of 0.896 between the measured and predicted carcinogenic potency with mean squared error (MSE) of 0.44 in complete rat data. The rat carcinogenicity model (GRNN) applied to the mouse and hamster data yielded correlation coefficient and MSE of 0.758, 0.71 and 0.760, 0.46, respectively. The results suggest for wide applicability of the inter-species models in predicting carcinogenic potency of chemicals. Both the PNN and GRNN (inter-species) models constructed here can be useful tools in predicting the carcinogenicity of new chemicals for regulatory purposes. - Graphical abstract: Figure (a) shows classification accuracies (positive and non-positive carcinogens) in rat, mouse, hamster, and pesticide data yielded by optimal PNN model. Figure (b) shows generalization and predictive abilities of the interspecies GRNN model to predict the carcinogenic potency of diverse chemicals. - Highlights: • Global robust models constructed for carcinogenicity prediction of diverse chemicals. • Tanimoto/BDS test revealed structural diversity of chemicals and nonlinearity in data. • PNN/GRNN successfully predicted carcinogenicity/carcinogenic potency of chemicals. • Developed interspecies PNN/GRNN models for carcinogenicity prediction. • Proposed models can be used as tool to predict carcinogenicity of new chemicals.
Forecasting of preprocessed daily solar radiation time series using neural networks
Paoli, Christophe; Muselli, Marc; Nivet, Marie-Laure [University of Corsica, CNRS UMR SPE, Corte (France); Voyant, Cyril [University of Corsica, CNRS UMR SPE, Corte (France); Hospital of Castelluccio, Radiotherapy Unit, Ajaccio (France)
2010-12-15
In this paper, we present an application of Artificial Neural Networks (ANNs) in the renewable energy domain. We particularly look at the Multi-Layer Perceptron (MLP) network which has been the most used of ANNs architectures both in the renewable energy domain and in the time series forecasting. We have used a MLP and an ad hoc time series pre-processing to develop a methodology for the daily prediction of global solar radiation on a horizontal surface. First results are promising with nRMSE {proportional_to} 21% and RMSE {proportional_to} 3.59 MJ/m{sup 2}. The optimized MLP presents predictions similar to or even better than conventional and reference methods such as ARIMA techniques, Bayesian inference, Markov chains and k-Nearest-Neighbors. Moreover we found that the data pre-processing approach proposed can reduce significantly forecasting errors of about 6% compared to conventional prediction methods such as Markov chains or Bayesian inference. The simulator proposed has been obtained using 19 years of available data from the meteorological station of Ajaccio (Corsica Island, France, 41 55'N, 8 44'E, 4 m above mean sea level). The predicted whole methodology has been validated on a 1.175 kWc mono-Si PV power grid. Six prediction methods (ANN, clear sky model, combination..) allow to predict the best daily DC PV power production at horizon d + 1. The cumulated DC PV energy on a 6-months period shows a great agreement between simulated and measured data (R{sup 2} > 0.99 and nRMSE < 2%). (author)
Forecasting Model for Crude Oil Price Using Artificial Neural Networks and Commodity Futures Prices
Kulkarni, Siddhivinayak
2009-01-01
This paper presents a model based on multilayer feedforward neural network to forecast crude oil spot price direction in the short-term, up to three days ahead. A great deal of attention was paid on finding the optimal ANN model structure. In addition, several methods of data pre-processing were tested. Our approach is to create a benchmark based on lagged value of pre-processed spot price, then add pre-processed futures prices for 1, 2, 3,and four months to maturity, one by one and also altogether. The results on the benchmark suggest that a dynamic model of 13 lags is the optimal to forecast spot price direction for the short-term. Further, the forecast accuracy of the direction of the market was 78%, 66%, and 53% for one, two, and three days in future conclusively. For all the experiments, that include futures data as an input, the results show that on the short-term, futures prices do hold new information on the spot price direction. The results obtained will generate comprehensive understanding of the cr...
Neural Networks with Non-Uniform Embedding and Explicit Validation Phase to Assess Granger Causality
Montalto, Alessandro; Faes, Luca; Tessitore, Giovanni; Prevete, Roberto; Marinazzo, Daniele
2015-01-01
A challenging problem when studying a dynamical system is to find the interdependencies among its individual components. Several algorithms have been proposed to detect directed dynamical influences between time series. Two of the most used approaches are a model-free one (transfer entropy) and a model-based one (Granger causality). Several pitfalls are related to the presence or absence of assumptions in modeling the relevant features of the data. We tried to overcome those pitfalls using a neural network approach in which a model is built without any a priori assumptions. In this sense this method can be seen as a bridge between model-free and model-based approaches. The experiments performed will show that the method presented in this work can detect the correct dynamical information flows occurring in a system of time series. Additionally we adopt a non-uniform embedding framework according to which only the past states that actually help the prediction are entered into the model, improving the prediction...
Combined expert system/neural networks method for process fault diagnosis
Reifman, Jaques (Westchester, IL); Wei, Thomas Y. C. (Downers Grove, IL)
1995-01-01
A two-level hierarchical approach for process fault diagnosis is an operating system employs a function-oriented approach at a first level and a component characteristic-oriented approach at a second level, where the decision-making procedure is structured in order of decreasing intelligence with increasing precision. At the first level, the diagnostic method is general and has knowledge of the overall process including a wide variety of plant transients and the functional behavior of the process components. An expert system classifies malfunctions by function to narrow the diagnostic focus to a particular set of possible faulty components that could be responsible for the detected functional misbehavior of the operating system. At the second level, the diagnostic method limits its scope to component malfunctions, using more detailed knowledge of component characteristics. Trained artificial neural networks are used to further narrow the diagnosis and to uniquely identify the faulty component by classifying the abnormal condition data as a failure of one of the hypothesized components through component characteristics. Once an anomaly is detected, the hierarchical structure is used to successively narrow the diagnostic focus from a function misbehavior, i.e., a function oriented approach, until the fault can be determined, i.e., a component characteristic-oriented approach.
Combined expert system/neural networks method for process fault diagnosis
Reifman, J.; Wei, T.Y.C.
1995-08-15
A two-level hierarchical approach for process fault diagnosis of an operating system employs a function-oriented approach at a first level and a component characteristic-oriented approach at a second level, where the decision-making procedure is structured in order of decreasing intelligence with increasing precision. At the first level, the diagnostic method is general and has knowledge of the overall process including a wide variety of plant transients and the functional behavior of the process components. An expert system classifies malfunctions by function to narrow the diagnostic focus to a particular set of possible faulty components that could be responsible for the detected functional misbehavior of the operating system. At the second level, the diagnostic method limits its scope to component malfunctions, using more detailed knowledge of component characteristics. Trained artificial neural networks are used to further narrow the diagnosis and to uniquely identify the faulty component by classifying the abnormal condition data as a failure of one of the hypothesized components through component characteristics. Once an anomaly is detected, the hierarchical structure is used to successively narrow the diagnostic focus from a function misbehavior, i.e., a function oriented approach, until the fault can be determined, i.e., a component characteristic-oriented approach. 9 figs.
Gu Zhanji; Ge Weigao [Department of Computer Science, Hainan Normal University, Haikou, HaiNan 571158 (China) and Department of Applied Mathematics, Beijing Institute of Technology, Beijing 100081 (China); Department of Applied Mathematics, Beijing Institute of Technology, Beijing 100081 (China)
2006-09-15
By using the continuation theorem of coincidence degree theory and constructing suitable Lyapunov functions, we study the existence, uniqueness, and global exponential stability of periodic solution for shunting inhibitory cellular neural networks with impulses, dx{sub ij}/dt=-a{sub ij}x{sub ij}-{sigma}{sub C{sub K}{sub 1}}{sub (set-membershipsign)=N{sub r}{sub (i,j)}C{sub ij}{sup kl}f{sub ij}[x{sub kl}(t)]x{sub ij}+L{sub ij}(t), t>0,t{ne}t{sub k}; {delta}x{sub ij}(t{sub k})=x{sub ij}(t{sub k}{sup +})-x{sub ij}(t{sub k}{sup -})=I{sub k}[x{sub ij}(t{sub k})], k=1,2,... . Furthermore, the numerical simulation shows that our system can occur in many forms of complexities, including periodic oscillation and chaotic strange attractor. To the best of our knowledge, these results have been obtained for the first time. Some researchers have introduced impulses into their models, but analogous results have never been found.
Lin, Tsau Young
control system of superheated steam temperature in power plants. The approach is simulated in MATLAB neural-network, TS model, Cascade control, Superheated steam temperature 1. Introduction There are two Network Based on the TS Model with Dynamic Consequent Parameters and its Application to Steam Temperature
Hagan, Martin
reactor, a robot arm, and a magnetic levitation system. 1. INTRODUCTION In this tutorial paper we want Model The multilayer perceptron neural network is built up of simple components. We will begin
Sen, Baris Ali; Menon, Suresh [School of Aerospace Engineering, Georgia Institute of Technology, 270 Ferst Drive, Atlanta, GA 30332-0150 (United States); Hawkes, Evatt R. [School of Photovoltaic and Renewable Energy Engineering, The University of New South Wales, NSW 2052 (Australia); School of Mechanical and Manufacturing Engineering, The University of New South Wales, NSW 2052 (Australia)
2010-03-15
Large eddy simulation (LES) of a non-premixed, temporally evolving, syngas/air flame is performed with special emphasis on speeding-up the sub-grid chemistry computations using an artificial neural networks (ANN) approach. The numerical setup for the LES is identical to a previous direct numerical simulation (DNS) study, which reported considerable local extinction and reignition physics, and hence, offers a challenging test case. The chemical kinetics modeling with ANN is based on a recent approach, and replaces the stiff ODE solver (DI) to predict the species reaction rates in the subgrid linear eddy mixing (LEM) model based LES (LEMLES). In order to provide a comprehensive evaluation of the current approach, additional information on conditional statistics of some of the key species and temperature are extracted from the previous DNS study and are compared with the LEMLES using ANN (ANN-LEMLES, hereafter). The results show that the current approach can detect the correct extinction and reignition physics with reasonable accuracy compared to the DNS. The syngas flame structure and the scalar dissipation rate statistics obtained by the current ANN-LEMLES are provided to further probe the flame physics. It is observed that, in contrast to H{sub 2}, CO exhibits a smooth variation within the region enclosed by the stoichiometric mixture fraction. The probability density functions (PDFs) of the scalar dissipation rates calculated based on the mixture fraction and CO demonstrate that the mean value of the PDF is insensitive to extinction and reignition. However, this is not the case for the scalar dissipation rate calculated by the OH mass fraction. Overall, ANN provides considerable computational speed-up and memory saving compared to DI, and can be used to investigate turbulent flames in a computationally affordable manner. (author)
{\\sc CosmoNet}: fast cosmological parameter estimation in non-flat models using neural networks
T. Auld; M. Bridges; M. P. Hobson
2007-03-16
We present a further development of a method for accelerating the calculation of CMB power spectra, matter power spectra and likelihood functions for use in cosmological Bayesian inference. The algorithm, called {\\sc CosmoNet}, is based on training a multilayer perceptron neural network. We compute CMB power spectra (up to $\\ell=2000$) and matter transfer functions over a hypercube in parameter space encompassing the $4\\sigma$ confidence region of a selection of CMB (WMAP + high resolution experiments) and large scale structure surveys (2dF and SDSS). We work in the framework of a generic 7 parameter non-flat cosmology. Additionally we use {\\sc CosmoNet} to compute the WMAP 3-year, 2dF and SDSS likelihoods over the same region. We find that the average error in the power spectra is typically well below cosmic variance for spectra, and experimental likelihoods calculated to within a fraction of a log unit. We demonstrate that marginalised posteriors generated with {\\sc CosmoNet} spectra agree to within a few percent of those generated by {\\sc CAMB} parallelised over 4 CPUs, but are obtained 2-3 times faster on just a \\emph{single} processor. Furthermore posteriors generated directly via {\\sc CosmoNet} likelihoods can be obtained in less than 30 minutes on a single processor, corresponding to a speed up of a factor of $\\sim 32$. We also demonstrate the capabilities of {\\sc CosmoNet} by extending the CMB power spectra and matter transfer function training to a more generic 10 parameter cosmological model, including tensor modes, a varying equation of state of dark energy and massive neutrinos. {\\sc CosmoNet} and interfaces to both {\\sc CosmoMC} and {\\sc Bayesys} are publically available at {\\tt www.mrao.cam.ac.uk/software/cosmonet}.
Nishitani, Y.; Kaneko, Y.; Ueda, M.; Fujii, E. [Advanced Technology Research Laboratories, Panasonic Corporation, Seika, Kyoto 619-0237 (Japan); Morie, T. [Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology, Wakamatsu-ku, Kitakyushu 808-0196 (Japan)
2012-06-15
Spike-timing-dependent synaptic plasticity (STDP) is demonstrated in a synapse device based on a ferroelectric-gate field-effect transistor (FeFET). STDP is a key of the learning functions observed in human brains, where the synaptic weight changes only depending on the spike timing of the pre- and post-neurons. The FeFET is composed of the stacked oxide materials with ZnO/Pr(Zr,Ti)O{sub 3} (PZT)/SrRuO{sub 3}. In the FeFET, the channel conductance can be altered depending on the density of electrons induced by the polarization of PZT film, which can be controlled by applying the gate voltage in a non-volatile manner. Applying a pulse gate voltage enables the multi-valued modulation of the conductance, which is expected to be caused by a change in PZT polarization. This variation depends on the height and the duration of the pulse gate voltage. Utilizing these characteristics, symmetric and asymmetric STDP learning functions are successfully implemented in the FeFET-based synapse device by applying the non-linear pulse gate voltage generated from a set of two pulses in a sampling circuit, in which the two pulses correspond to the spikes from the pre- and post-neurons. The three-terminal structure of the synapse device enables the concurrent learning, in which the weight update can be performed without canceling signal transmission among neurons, while the neural networks using the previously reported two-terminal synapse devices need to stop signal transmission for learning.
A nanoflare model for active region radiance: application of artificial neural networks
M. Bazarghan; H. Safari; D. E. Innes; E. Karami; S. K. Solanki
2008-12-20
Context. Nanoflares are small impulsive bursts of energy that blend with and possibly make up much of the solar background emission. Determining their frequency and energy input is central to understanding the heating of the solar corona. One method is to extrapolate the energy frequency distribution of larger individually observed flares to lower energies. Only if the power law exponent is greater than 2, is it considered possible that nanoflares contribute significantly to the energy input. Aims. Time sequences of ultraviolet line radiances observed in the corona of an active region are modelled with the aim of determining the power law exponent of the nanoflare energy distribution. Methods. A simple nanoflare model based on three key parameters (the flare rate, the flare duration time, and the power law exponent of the flare energy frequency distribution) is used to simulate emission line radiances from the ions Fe XIX, Ca XIII, and Si iii, observed by SUMER in the corona of an active region as it rotates around the east limb of the Sun. Light curve pattern recognition by an Artificial Neural Network (ANN) scheme is used to determine the values. Results. The power law exponents, alpha 2.8, 2.8, and 2.6 for Fe XIX, Ca XIII, and Si iii respectively. Conclusions. The light curve simulations imply a power law exponent greater than the critical value of 2 for all ion species. This implies that if the energy of flare-like events is extrapolated to low energies, nanoflares could provide a significant contribution to the heating of active region coronae.
Behrang, M.A.; Assareh, E. [Department of Mechanical Engineering, Young Researchers Club, Islamic Azad University, Dezful Branch (Iran); Ghanbarzadeh, A.; Noghrehabadi, A.R. [Department of Mechanical Engineering, Engineering Faculty, Shahid Chamran University, Ahvaz (Iran)
2010-08-15
The main objective of present study is to predict daily global solar radiation (GSR) on a horizontal surface, based on meteorological variables, using different artificial neural network (ANN) techniques. Daily mean air temperature, relative humidity, sunshine hours, evaporation, and wind speed values between 2002 and 2006 for Dezful city in Iran (32 16'N, 48 25'E), are used in this study. In order to consider the effect of each meteorological variable on daily GSR prediction, six following combinations of input variables are considered: (I)Day of the year, daily mean air temperature and relative humidity as inputs and daily GSR as output. (II)Day of the year, daily mean air temperature and sunshine hours as inputs and daily GSR as output. (III)Day of the year, daily mean air temperature, relative humidity and sunshine hours as inputs and daily GSR as output. (IV)Day of the year, daily mean air temperature, relative humidity, sunshine hours and evaporation as inputs and daily GSR as output. (V)Day of the year, daily mean air temperature, relative humidity, sunshine hours and wind speed as inputs and daily GSR as output. (VI)Day of the year, daily mean air temperature, relative humidity, sunshine hours, evaporation and wind speed as inputs and daily GSR as output. Multi-layer perceptron (MLP) and radial basis function (RBF) neural networks are applied for daily GSR modeling based on six proposed combinations. The measured data between 2002 and 2005 are used to train the neural networks while the data for 214 days from 2006 are used as testing data. The comparison of obtained results from ANNs and different conventional GSR prediction (CGSRP) models shows very good improvements (i.e. the predicted values of best ANN model (MLP-V) has a mean absolute percentage error (MAPE) about 5.21% versus 10.02% for best CGSRP model (CGSRP 5)). (author)
Bodruzzaman, M.; Essawy, M.A.
1996-02-27
Pressurized fluidized-bed combustors (FBC) are becoming very popular, efficient, and environmentally acceptable replica for conventional boilers in Coal-fired and chemical plants. In this paper, we present neural network-based methods for chaotic behavior monitoring and control in FBC systems, in addition to chaos analysis of FBC data, in order to localize chaotic modes in them. Both of the normal and abnormal mixing processes in FBC systems are known to undergo chaotic behavior. Even though, this type of behavior is not always undesirable, it is a challenge to most types of conventional control methods, due to its unpredictable nature. The performance, reliability, availability and operating cost of an FBC system will be significantly improved, if an appropriate control method is available to control its abnormal operation and switch it to normal when exists. Since this abnormal operation develops only at certain times due to a sequence of transient behavior, then an appropriate abnormal behavior monitoring method is also necessary. Those methods has to be fast enough for on-line operation, such that the control methods would be applied before the system reaches a non-return point in its transients. It was found that both normal and abnormal behavior of FBC systems are chaotic. However, the abnormal behavior has a higher order chaos. Hence, the appropriate control system should be capable of switching the system behavior from its high order chaos condition to low order chaos. It is to mention that most conventional chaos control methods are designed to switch a chaotic behavior to a periodic orbit. Since this is not the goal for the FBC case, further developments are needed. We propose neural network-based control methods which are known for their flexibility and capability to control both non-linear and chaotic systems. A special type of recurrent neural network, known as Dynamic System Imitator (DSI), will be used for the monitoring and control purposes.
An Approximation to the Cross Sections of Z_l Boson Production at CLIC by Using Neural Networks
S. Akkoyun; S. O. Kara
2012-12-14
In this work, the possible dynamics associated with leptophilic Z_l boson at CLIC (Compact Linear Collider) have been investigated by using artificial neural networks (ANNs). These hypotetic massive boson Z_l have been shown through the process e+e- -> M+M-. Furthermore, the invariant mass distributions for final muons have been consistently predicted by using ANN. For these highly non-linear data, we have constructed consistent empirical physical formulas (EPFs) by appropriate feed- forward ANN. These ANN-EPFs can be used to derive further physical functions which could be relevant to studying Z_l.
Keller, Paul E.; McMakin, Douglas L.; Hall, Thomas E.; Sheen, David M.
2006-06-01
Events in the past few years have heightened security concerns necessitating the development of more advanced methods for detecting potential threats being carried on individuals. One approach is to use imaging methods that see through clothing to find potentially threatening objects being concealed by individuals on their person. This sparks obvious privacy concerns. This paper describes one technique based on neural networks and Fourier features applied to active millimeter-wave imagery that finds man-made structure potentially indicating a threat without compromising personal privacy.
Teimoorinia, H.
2012-12-01
The aim of this work is to combine spectral energy distribution (SED) fitting with artificial neural network techniques to assign spectral characteristics to a sample of galaxies at 0.5 < z < 1. The sample is selected from the spectroscopic campaign of the ESO/GOODS-South field, with 142 sources having photometric data from the GOODS-MUSIC catalog covering bands between {approx}0.4 and 24 {mu}m in 10-13 filters. We use the CIGALE code to fit photometric data to Maraston's synthesis spectra to derive mass, specific star formation rate, and age, as well as the best SED of the galaxies. We use the spectral models presented by Kinney et al. as targets in the wavelength interval {approx}1200-7500 A. Then a series of neural networks are trained, with average performance {approx}90%, to classify the best SED in a supervised manner. We consider the effects of the prominent features of the best SED on the performance of the trained networks and also test networks on the galaxy spectra of Coleman et al., which have a lower resolution than the target models. In this way, we conclude that the trained networks take into account all the features of the spectra simultaneously. Using the method, 105 out of 142 galaxies of the sample are classified with high significance. The locus of the classified galaxies in the three graphs of the physical parameters of mass, age, and specific star formation rate appears consistent with the morphological characteristics of the galaxies.
Avignon et des Pays de Vaucluse, UniversitÃ© de
ESANN'1995 proceedings - European Symposium on Artificial Neural Networks Brussels (Belgium), 19-20-21 April 1995, D-Facto public., ISBN 2-9600049-3-0, pp. 365-371 #12;ESANN'1995 proceedings - European Symposium on Artificial Neural Networks Brussels (Belgium), 19-20-21 April 1995, D-Facto public., ISBN 2
Verleysen, Michel
ESANN'1995 proceedings - European Symposium on Artificial Neural Networks Brussels (Belgium), 19-20-21 April 1995, D-Facto public., ISBN 2-9600049-3-0, pp. 153-160 #12;ESANN'1995 proceedings - European Symposium on Artificial Neural Networks Brussels (Belgium), 19-20-21 April 1995, D-Facto public., ISBN 2
Li, Jun; Guo, Hua E-mail: hguo@unm.edu; Chen, Jun; Zhang, Dong H. E-mail: hguo@unm.edu
2014-01-28
A permutationally invariant global potential energy surface for the HOCO system is reported by fitting a larger number of high-level ab initio points using the newly proposed permutation invariant polynomial-neural network method. The small fitting error (?5 meV) indicates a faithful representation of the potential energy surface over a large configuration space. Full-dimensional quantum and quasi-classical trajectory studies of the title reaction were performed on this potential energy surface. While the results suggest that the differences between this and an earlier neural network fits are small, discrepancies with state-to-state experimental data remain significant.
Voyant, Cyril; Paoli, Christophe; Nivet, Marie Laure; Poggi, Philippe
2009-01-01
The official meteorological network is poor on the island of Corsica: only three sites being about 50 km apart are equipped with pyranometers which enable measurements by hourly and daily step. These sites are Ajaccio (seaside), Bastia (seaside) and Corte (average altitude of 486 meters). This lack of weather station makes difficult the predictability of PV power grid performance. This work intends to study a methodology which can predict global solar irradiation using data available from another location for daily and hourly horizon. In order to achieve this prediction, we have used Artificial Neural Network which is a popular artificial intelligence technique in the forecasting domain. A simulator has been obtained using data available for the station of Ajaccio that is the only station for which we have a lot of data: 16 years from 1972 to 1987. Then we have tested the efficiency of this simulator in two places with different geographical features: Corte, a mountainous region and Bastia, a coastal region. ...
An Adaptive Landscape Classification Procedure using Geoinformatics and Artificial Neural Networks
Coleman, Andre M.
2008-08-01
The Adaptive Landscape Classification Procedure (ALCP), which links the advanced geospatial analysis capabilities of Geographic Information Systems (GISs) and Artificial Neural Networks (ANNs) and particularly Self-Organizing Maps (SOMs), is proposed as a method for establishing and reducing complex data relationships. Its adaptive and evolutionary capability is evaluated for situations where varying types of data can be combined to address different prediction and/or management needs such as hydrologic response, water quality, aquatic habitat, groundwater recharge, land use, instrumentation placement, and forecast scenarios. The research presented here documents and presents favorable results of a procedure that aims to be a powerful and flexible spatial data classifier that fuses the strengths of geoinformatics and the intelligence of SOMs to provide data patterns and spatial information for environmental managers and researchers. This research shows how evaluation and analysis of spatial and/or temporal patterns in the landscape can provide insight into complex ecological, hydrological, climatic, and other natural and anthropogenic-influenced processes. Certainly, environmental management and research within heterogeneous watersheds provide challenges for consistent evaluation and understanding of system functions. For instance, watersheds over a range of scales are likely to exhibit varying levels of diversity in their characteristics of climate, hydrology, physiography, ecology, and anthropogenic influence. Furthermore, it has become evident that understanding and analyzing these diverse systems can be difficult not only because of varying natural characteristics, but also because of the availability, quality, and variability of spatial and temporal data. Developments in geospatial technologies, however, are providing a wide range of relevant data, and in many cases, at a high temporal and spatial resolution. Such data resources can take the form of high-dimensional data arrays, which can difficult to fully use. Establishing relationships among high-dimensional datasets through neurocomputing based patterning methods can help 1) resolve large volumes of data into a meaningful form; 2) provide an approach for inferring landscape processes in areas that have limited data available but that exhibit similar landscape characteristics; and 3) discover the value of individual variables or groups of variables that contribute to specific processes in the landscape.
Zhao, W [Triton Systems, Inc.; Pinnaduwage, Lal A [ORNL; Leis, J. W. [University of Southern Queensland; Gehl, Anthony C [ORNL; Allman, Steve L [ORNL; Shepp, A. [Triton Systems, Inc.; Mahmud, K. [Triton Systems, Inc.
2008-01-01
We report the experimental details on the successful application of the electronic nose approach to identify and quantify components in ternary vapor mixtures. Preliminary results have recently been presented [L. A. Pinnaduwage et al., Appl. Phys. Lett. 91, 044105 (2007)]. Our microelectromechanical-system-based electronic nose is composed of a microcantilever sensor array with seven individual sensors used for vapor detection and an artificial neural network for pattern recognition. A set of custom vapor generators generated reproducible vapor mixtures in different compositions for training and testing of the neural network. The sensor array was selected to be capable of generating different response patterns to mixtures with different component proportions. Therefore, once the electronic nose was trained by using the response patterns to various compositions of the mixture, it was able to predict the composition of 'unknown' mixtures. We have studied two vapor systems: one included the nerve gas simulant dimethylmethyl phosphonate at ppb concentrations and water and ethanol at ppm concentrations; the other system included acetone, water, and ethanol all of which were at ppm concentrations. In both systems, individual, binary, and ternary mixtures were analyzed with good reproducibility.
Bourouis, Chahrazed [Faculty of science and Engineering University of Guelma, BP 401 Guelma 24000 (Algeria); Meddour, Ahcene [Laboratory of semi conductors, University of Badji Mokhtar, Annaba (Algeria); Moussaoui, Abdelkrim [Electrical Engineering Laboratory (LGEG), University of Guelma, BP 401, 24000 (Algeria)
2008-09-23
In this paper a new method using the combination of Neural Networks and the Newton-Raphson algorithm is developped. The technique consists of the use of the solution obtained by Newton-Raphson algorithm between 0.5 and 2.1eV for pure manganese (Mn) and for the amorphous metallic alloy Al{sub 88}Mn{sub 12}, to construct two parts of datasets; the first one is used for training the neural network and the second one for the validation tests. The validated neural network model is applied to the determination of optical constants of the two materials Mn and Al{sub 88}Mn{sub 12} in the range of 0.5 and 6.2eV (IR-VIS-UV). The results obtained over all the studied energy range are used to trace back to dielectric function, optical absorption and electronic structure of the same material. By using the partial solution obtained by Newton-Raphson as a database of the neural network prediction model, it is shown that the obtained results are in accordance with those of the literature which consolidate the efficiency of the suggested approach.
Triantaphyllou, Evangelos
Detection of Welding Flaws with MLP Neural Network and Case Based Reasoning T. Warren Liao1 *, E-Ze University, Nei-Li 32026, Chung-Li, Taiwan Abstract - The correct detection of welding flaws is important to the successful development of an automated weld inspection system. As a continuation of our previous efforts
Barranco, Bernabe Linares
1196 IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 19, NO. 7, JULY 2008 On Real-Time AER 2-Devent-rep- resentation (AER) technique, which is a spike-based biologically inspired image and video representation interfaces have been developed for generating AER streams from conven- tional computers and feeding them
Simões, Marcelo Godoy
of the dynamic behavior of a system consisting of a turret-floating production storage and offloading (FPSO) system and a shuttle ship in tandem configuration is de- scribed. The turret-FPSO is a vessel to be obtained. In order to deal with such complexities, a neural network has been devised to simulate an FPSO
Paris-Sud XI, Université de
PRODUCTION PLANNING AND CONTROL, VOL.22 N°8 PP 767-781 2011 A neural network for the reduction manuscript, published in "Production Planning & Control 22, 8 (2011) 767-781" DOI : 10 manufacturing systems, planning and control processes simulation is essential for evaluation of planning
Lehman, Brad
on a shaded solar panel at different hours of a day for several days. After training the neural network, its, building-integrated photovoltaic panels, and portable solar tents, it is common for a solar PV to become output power of a solar photovoltaic array under the non-uniform shadow conditions at a given geographic
Neal, Radford M.
, you will #12;t Bayesian neural network models with several architectures to data from a project I am of a complex simulation of glaciation in North America over the last 250,000 years. The aim of the project over this time period, etc. These parameters need to be estimated on the basis of how well
Neal, Radford M.
, you will fit Bayesian neural network models with several architectures to data from a project I am of a complex simulation of glaciation in North America over the last 250,000 years. The aim of the project over this time period, etc. These parameters need to be estimated on the basis of how well
Corani, Giorgio
2005-01-01
Ecological Modelling 185 (2005) 513529 Air quality prediction in Milan: feed-forward neural December 2004; accepted 3 January 2005 Abstract Ozone and PM10 constitute the major concern for air quality of Milan. This paper addresses the problem of the prediction of such two pollutants, using to this end
3-D Inversion Of Borehole-To-Surface Electrical Data Using A...
3-D Inversion Of Borehole-To-Surface Electrical Data Using A Back-Propagation Neural Network Jump to: navigation, search OpenEI Reference LibraryAdd to library Journal Article: 3-D...
Neural-network selection of high-redshift radio quasars, and the luminosity function at z~4
Tuccillo, D; Benn, C R
2015-01-01
We obtain a sample of 87 radio-loud QSOs in the redshift range 3.6 1 mJy with star-like objects having r neural-network, which yields 97% completeness (fraction of actual high-z QSOs selected as such) and an efficiency (fraction of candidates which are high-z QSOs) in the range of 47 to 60%. We use this sample to estimate the binned optical luminosity function of radio-loud QSOs at $z\\sim 4$, and also the LF of the total QSO population and its comoving density. Our results suggest that the radio-loud fraction (RLF) at high z is similar to that at low-z and that other authors may be underestimating the fraction at high-z. Finally, we...
López-Caraballo, C H; Salfate, I; Rojas, P; Rivera, M; Palma-Chilla, L
2015-01-01
In this study, an artificial neural network (ANN) based on particle swarm optimization (PSO) was developed for the time series prediction. The hybrid ANN+PSO algorithm was applied on Mackey--Glass chaotic time series in the short-term $x(t+6)$. The performance prediction was evaluated and compared with another studies available in the literature. Also, we presented properties of the dynamical system via the study of chaotic behaviour obtained from the predicted time series. Next, the hybrid ANN+PSO algorithm was complemented with a Gaussian stochastic procedure (called {\\it stochastic} hybrid ANN+PSO) in order to obtain a new estimator of the predictions, which also allowed us to compute uncertainties of predictions for noisy Mackey--Glass chaotic time series. Thus, we studied the impact of noise for several cases with a white noise level ($\\sigma_{N}$) from 0.01 to 0.1.
Tolbert, Leon M; Ozpineci, Burak; Filho, Faete; Cao, Yue
2011-01-01
This work approximates the selective harmonic elimination problem using artificial neural networks (ANNs) to generate the switching angles in an 11-level full-bridge cascade inverter powered by five varying dc input sources. Each of the five full bridges of the cascade inverter was connected to a separate 195-W solar panel. The angles were chosen such that the fundamental was kept constant and the low-order harmonics were minimized or eliminated. A nondeterministic method is used to solve the system for the angles and to obtain the data set for the ANN training. The method also provides a set of acceptable solutions in the space where solutions do not exist by analytical methods. The trained ANN is a suitable tool that brings a small generalization effect on the angles' precision and is able to perform in real time (50-/60-Hz time window).
Zwink, AB; Turner, DD
2012-03-19
The fore-optics of the Atmospheric Emitted Radiance Interferometer (AERI) are protected by an automated hatch to prevent precipitation from fouling the instrument's scene mirror (Knuteson et al. 2004). Limit switches connected with the hatch controller provide a signal of the hatch state: open, closed, undetermined (typically associated with the hatch being between fully open or fully closed during the instrument's sky view period), or an error condition. The instrument then records the state of the hatch with the radiance data so that samples taken when the hatch is not open can be removed from any subsequent analysis. However, the hatch controller suffered a multi-year failure for the AERI located at the ARM North Slope of Alaska (NSA) Central Facility in Barrow, Alaska, from July 2006-February 2008. The failure resulted in misreporting the state of the hatch in the 'hatchOpen' field within the AERI data files. With this error there is no simple solution to translate what was reported back to the correct hatch status, thereby making it difficult for an analysis to determine when the AERI was actually viewing the sky. As only the data collected when the hatch is fully open are scientifically useful, an algorithm was developed to determine whether the hatch was open or closed based on spectral radiance data from the AERI. Determining if the hatch is open or closed in a scene with low clouds is non-trivial, as low opaque clouds may look very similar spectrally as the closed hatch. This algorithm used a backpropagation neural network; these types of neural networks have been used with increasing frequency in atmospheric science applications.
Cho, Daniel D; Wernicke, A Gabriella; Nori, Dattatreyudu; Chao, KSC; Parashar, Bhupesh; Chang, Jenghwa [Weill Cornell Medical College, NY, NY (United States)
2014-06-01
Purpose/Objective(s): The aim of this study is to build the estimator of toxicity using artificial neural network (ANN) for head and neck cancer patients Materials/Methods: An ANN can combine variables into a predictive model during training and considered all possible correlations of variables. We constructed an ANN based on the data from 73 patients with advanced H and N cancer treated with external beam radiotherapy and/or chemotherapy at our institution. For the toxicity estimator we defined input data including age, sex, site, stage, pathology, status of chemo, technique of external beam radiation therapy (EBRT), length of treatment, dose of EBRT, status of post operation, length of follow-up, the status of local recurrences and distant metastasis. These data were digitized based on the significance and fed to the ANN as input nodes. We used 20 hidden nodes (for the 13 input nodes) to take care of the correlations of input nodes. For training ANN, we divided data into three subsets such as training set, validation set and test set. Finally, we built the estimator for the toxicity from ANN output. Results: We used 13 input variables including the status of local recurrences and distant metastasis and 20 hidden nodes for correlations. 59 patients for training set, 7 patients for validation set and 7 patients for test set and fed the inputs to Matlab neural network fitting tool. We trained the data within 15% of errors of outcome. In the end we have the toxicity estimation with 74% of accuracy. Conclusion: We proved in principle that ANN can be a very useful tool for predicting the RT outcomes for high risk H and N patients. Currently we are improving the results using cross validation.
Short-Term Load Forecasting by Feed-Forward Neural Networks Saied S. Sharif1
Taylor, James H.
, Department of Electrical and Computer Engineering, University of New Brunswick, Fredericton, NB, CANADA, E3B) is presented for the hourly load forecasting of the coming days. In this approach, 24 independent networks are used for the next day load forecast. Each network is utilized for the prediction of load at a specific
Voyant, Cyril; Paoli, Christophe; Nivet, Marie Laure; Poggi, Philippe; Haurant, P; 10.4229/24thEUPVSEC2009-5BV.2.35
2010-01-01
The official meteorological network is poor on the island of Corsica: only three sites being about 50 km apart are equipped with pyranometers which enable measurements by hourly and daily step. These sites are Ajaccio (41\\degree 55'N and 8\\degree 48'E, seaside), Bastia (42\\degree 33'N, 9\\degree 29'E, seaside) and Corte (42\\degree 30'N, 9\\degree 15'E average altitude of 486 meters). This lack of weather station makes difficult the predictability of PV power grid performance. This work intends to study a methodology which can predict global solar irradiation using data available from another location for daily and hourly horizon. In order to achieve this prediction, we have used Artificial Neural Network which is a popular artificial intelligence technique in the forecasting domain. A simulator has been obtained using data available for the station of Ajaccio that is the only station for which we have a lot of data: 16 years from 1972 to 1987. Then we have tested the efficiency of this simulator in two places w...
Impact of synaptic depression on network activity and implications for neural coding
York, Lawrence Christopher
2011-11-24
Short-term synaptic depression is the phenomena where repeated stimulation leads to a decreased transmission efficacy. In this thesis, the impact of synaptic depression on the responses and dynamics of network models of ...
A Training Scheme for Pattern Classification Using Multi-layer Feed-forward Neural Networks.
Keeni, Kanad; Nakayama, Kenji; Shimodaira, Hiroshi
1999-01-01
This study highlights on the subject of weight initialization in multi-layer feed-forward networks. Training data is analyzed and the notion of criti- cal point is introduced for determining the initial weights for the ...
Fowler, M.M.; Klett, D.E.; Moreno, J.B.; Heermann, P.D.
1997-03-01
Liquid metal reflux receivers (LMRRs) have been designed to serve as the interface between the solar concentrator dish and the Stirling engine of a dish Stirling power system. Such a receiver has undergone performance testing at Sandia National Laboratory to determine cold- and hot-start characteristics, component temperatures, throughput power, and thermal efficiency, for various times of day and year. Performance modeling will play an important role in the future commercialization of these systems since it will be necessary to predict overall energy production for potential installation sites based on available meteorological data. As a supplement to numerical thermal modeling, artificial neural networks (ANNs) have been investigated for their effectiveness in predicting long-term energy production of a LMRR. Two types of data were used to train ANNs, actual on-sun test data, and ersatz data. ANNs were trained on both the raw on-sun test data and on pre-formatted versions of the data to determine if pre-formatting of the input data would improve network training efficiency and predictive abilities. Usable on-sun test data were available for only a few days of performance testing. Therefore, a set of year-long ersatz data was generated using a transient numerical model driven by one-minute meteorological data from the Solar Energy Meteorological Research and Training Sites (SEMRTS) data base for Davis, CA. The ersatz data were used to train ANNs based on warm-month data, cool-month data, and year-long data to investigate the impact of using seasonal test data on long-term predictive capabilities. The findings indicated that a network trained on data from a limited time span could successfully predict annual energy output of a liquid metal receiver.
Kirwan, Peter; Turner-Bridger, Benita; Peter, Manuel; Momoh, Ayiba; Arambepola, Devika; Robinson, Hugh P.C.; Livesey, Frederick J.
2015-07-30
., 2012). Computational modelling of connectivity in the cortex suggests that networks adhere to a scale-free topology, which include small numbers of hyper-connected hub-like nodes, and a large number of neurons with few connections (Eguiluz et al., 2005... contribute to the observed network properties. Using rabies virus-based trans-synaptic tracing (Wickersham et al., 2007), we found that the majority of cortical neurons in this system receive inputs from fewer than 10 neurons, with a small minority receiving...