Artificial Bee Colony Training of Neural Networks: Comparison with Back-Propagation
Bullinaria, John
Artificial Bee Colony Training of Neural Networks: Comparison with Back-Propagation John A: Artificial Bee Colony, Neural Networks, Learning, Evolution. A copy-edited version of this paper has been for optimization that has previously been applied to the training of neural networks. This paper examines more
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
1997-01-01T23:59:59.000Z
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
Training a 3-Node Neural Network is NP-Complete
Rivest, Ronald L.
. A good discussion of the theory of NP-completeness, as well as a description of several hundreds of NPTraining a 3-Node Neural Network is NP-Complete Avrim L. Blum and Ronald L. Rivest* MIT Laboratory-propagation algorithm promises just that. In practice, however, the back-propagation algorithm often runs very slowly
Automatic real-time lip synchronization using LPC analysis and neural networks
Haaser, Christina Marie
2002-01-01T23:59:59.000Z
, specifically LPC cepstral coefficients, gain, and zero-crossing rate. These features are used as input into a trained three-layer feedforward back-propagation neural network for phonetic classification per frame of animation. The training of the neural network...
Computationally Efficient Neural Network Intrusion Security Awareness
Todd Vollmer; Milos Manic
2009-08-01T23:59:59.000Z
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.
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 by the back- propagation algorithm. A set of load profiles as inputs and the corresponding unit-commitment
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
Kjellström, Hedvig
Artificial Neural Networks Single Layer Networks Multi Layer Networks Generalization Artificial Neural Networks Artificial Neural Networks Single Layer Networks Multi Layer Networks Generalization 1 Artificial Neural Networks Properties Applications Classical Examples Biological Background 2 Single Layer
Kjellström, Hedvig
Artificial Neural Networks Single Layer Networks Multi Layer Networks Generalization Artificial Neural Networks #12;Artificial Neural Networks Single Layer Networks Multi Layer Networks Generalization 1 Artificial Neural Networks Properties Applications Classical Examples Biological Background 2
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))
1991-01-01T23:59:59.000Z
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.
Gabel, S.
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...
Gabel, S.
2003-01-01T23:59:59.000Z
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...
Jeffery, Simon
1 and J.S. Drilling 2 1 Armagh Observatory, College Hill, Armagh BT61 9DG, N. Ireland 2 Dept@arm.ac.uk, drilling@rouge.phys.lsu.edu Abstract We apply a multilayer feed-forward back propagation arti#12;cial neural network to a sample of 380 subdwarf spectra classi#12;ed by Drilling et al. (2002), showing
Misbehavior in a Neural Network Model
Burgos, José E
2015-01-01T23:59:59.000Z
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
2015-01-01T23:59:59.000Z
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
Foundations of Artificial Intelligence Neural Networks
Qu, Rong
Foundations of Artificial Intelligence Neural Networks Building Artificial Brains #12;Background Session 2 Software Demonstrations Real World Applications #12;Artificial Neural Networks ... consists to operate. Wiki #12;Relationship between Artificial Neural Networks & the Human Brain Neural networks
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
2009-07-01T23:59:59.000Z
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.
Bayesian and Maximum Likelihood Neural Networks
Nielsen, Finn Ã?rup
Bayesian and Maximum Likelihood Neural Networks Finn A ffi rup Nielsen Section for Digital Signal, 1998 OVERVIEW ffl Artificial neural networks ffl Maximum likelihood, MAP, MPL neural networks ffl Bayesian neural networks -- MCMC Bayesian neural networks \\Lambda Hybrid Monte Carlo ffl lyngby matlab
A VALIDATION INDEX FOR ARTIFICIAL NEURAL NETWORKS
Roberts, Stephen
A VALIDATION INDEX FOR ARTIFICIAL NEURAL NETWORKS Stephen Roberts, Lionel Tarassenko, James Pardey and estimation properties of artificial neural networks. Like many `traditional' statistical techniques & David Siegwart Neural Network Research Group Department of Engineering Science University of Oxford, UK
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)
1996-08-01T23:59:59.000Z
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.
Development of Fast-Running Simulation Methodology Using Neural Networks for Load Follow Operation
Seong, Seung-Hwan [Korea Atomic Energy Research Institute (Korea, Republic of); Park, Heui-Youn [Korea Atomic Energy Research Institute (Korea, Republic of); Kim, Dong-Hoon [Korea Atomic Energy Research Institute (Korea, Republic of); Suh, Yong-Suk [Korea Atomic Energy Research Institute (Korea, Republic of); Hur, Seop [Korea Atomic Energy Research Institute (Korea, Republic of); Koo, In-Soo [Korea Atomic Energy Research Institute (Korea, Republic of); Lee, Un-Chul [Seoul National University (Korea, Republic of); Jang, Jin-Wook [Seoul National University (Korea, Republic of); Shin, Yong-Chul [Yonsei University (Korea, Republic of)
2002-05-15T23:59:59.000Z
A new fast-running analytic model has been developed for analyzing the load follow operation. The new model was based on the neural network theory, which has the capability of modeling the input/output relationships of a nonlinear system. The new model is made up of two error back-propagation neural networks and procedures to calculate core parameters, such as the distributions and density of xenon in a quasi-steady-state core like load follow operation. One neural network is designed to retrieve the axial offset of power distribution, and the other is for reactivity corresponding to a given core condition. The training data sets for learning the neural networks in the new model are generated with a three-dimensional nodal code and, also, the measured data of the first-day test of load follow operation. Using the new model, the simulation results of the 5-day load follow test in a pressurized water reactor show a good agreement between the simulation data and the actual measured data. Required computing time for simulating a load follow operation is comparable to that of a fast-running lumped model. Moreover, the new model does not require additional engineering factors to compensate for the difference between the actual measurements and analysis results because the neural network has the inherent learning capability of neural networks to new situations.
Calibrating Artificial Neural Networks by Global Optimization
Janos D. Pinter
2010-07-21T23:59:59.000Z
Jul 21, 2010 ... Abstract: An artificial neural network (ANN) is a computational model ... emulating the key features and operations of biological neural networks.
Back Propagation is Sensitive to Initial Conditions John F. Kolen
Pollack, Jordan B.
Back Propagation is Sensitive to Initial Conditions John F. Kolen Jordan B. Pollack Laboratory Columbus, Ohio 43210, USA kolenj@cis.ohiostate.edu, pollack@cis.ohiostate.edu TR 90JKBPSIC ABSTRACT. Kolen Jordan B. Pollack Laboratory for Artificial Intelligence Research Computer and Information Science
Back Propagation is Sensitive to Initial Conditions John F. Kolen
Pollack, Jordan B.
Back Propagation is Sensitive to Initial Conditions John F. Kolen Jordan B. Pollack Laboratory Columbus, Ohio 43210, USA kolen-j@cis.ohio-state.edu, pollack@cis.ohio-state.edu TR 90-JK-BPSIC ABSTRACT. Kolen Jordan B. Pollack Laboratory for Artificial Intelligence Research Computer and Information Science
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
Semiring Artificial Neural Networks and Weighted Automata
Hoelldobler, Steffen
Semiring Artificial Neural Networks and Weighted Automata And an Application to Digital Image neural networks and weighted automata. For this task, we introduce semiring artificial neural networks, that is, artificial neural networks which implement the addition and the multiplication of semirings. We
Artificial Neural Network Portion of Coil Study
Putten, Peter van der
Artificial Neural Network Portion of Coil Study LTC William M. Crocoll School of Systems TO ARTIFICIAL NEURAL NETWORKS A neural network is a massively parallel system comprised of many highly of the brain (Dayhoff, 1990). A major task for a neural network is to learn and maintain a model of the world
Artificial Bee Colony Training of Neural Networks
Bullinaria, John
Artificial Bee Colony Training of Neural Networks John A. Bullinaria and Khulood AlYahya School of artificial Neural Networks (NNs). Of course, there already exist many hybrid neural network learning for optimization, that has previously been applied successfully to the training of neural networks. This paper ex
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
2015-01-01T23:59:59.000Z
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
2009-01-01T23:59:59.000Z
2.3 Artificial Neural Networks . . . . . . . . . . .2.3.1 Learning in Artificial Neural Networks 2.3.2 HopfieldBayesian learning for neural networks. Springer Verlag,
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
2006-01-01T23:59:59.000Z
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
Solos (Dice Game) and Conductor (Neural Network)
Marquetti, Andre
2015-01-01T23:59:59.000Z
method to compose music and the Conductors’ neural network for processing the music is based on pattern- extraction,
Object Oriented Artificial Neural Network Implementations
Slatton, Clint
1 Object Oriented Artificial Neural Network Implementations W. Curt Lefebvre Jose C. Principe Neuro artificial neural networks (ANNs). The conven- tion for ANN simulation has been a direct implementation to develop a graphical artificial neural network simulation environment motivated towards the pro- cessing
Online learning processes artificial neural networks
Heskes, Tom
Online learning processes in artificial neural networks Tom M. Heskes Bert Kappen Department, The Netherlands. Abstract We study online learning processes in artificial neural networks from a general point. Elsevier, pages 199-- 233. #12; Online learning processes in artificial neural networks 1 1 Introduction 1
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
Delaware, University of
1 1 Introduction The artificial neural network discussed in this paper is unlike most other neural practically all types of artificial neural network architectures were modeled, with varying levels of Connection Patterns for a Dynamic Artificial Neural Network Proceedings of the Combinations of Genetic
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)
2013-02-15T23:59:59.000Z
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.
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
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)
2011-01-17T23:59:59.000Z
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.
Neural networks for fast image compression
Li, Mu
1998-01-01T23:59:59.000Z
of one network have been used in the system. Each neural network was trained with some image blocks which have similar characteristics. In order to decrease the time for the learning process of the neural networks to converge, an adaptive back...
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)
1998-01-01T23:59:59.000Z
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.
1998-04-28T23:59:59.000Z
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.
Algorithms and Hardware for Implementing Artificial Neural Networks Nathan Hower
Algorithms and Hardware for Implementing Artificial Neural Networks Nathan Hower Abstract Complex problems require sophisticated processing techniques. Artificial neural networks are based require a parallel processing approach to be computed at practical speeds. Artificial neural networks
Nonlinear adaptive internal model control using neural networks
Gandhi, Amit Krushnavadan
2001-01-01T23:59:59.000Z
(NIMC) strategy based on neural network models is presented for SISO processes. The nonlinearities of the dynamical system are modelled by neural network architectures. Recurrent neural networks can be used for both the identification and control of nonlinear...
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
2013-01-01T23:59:59.000Z
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
The neural network approach to parton distributions
Andrea Piccione; Joan Rojo; for the NNPDF Collaboration
2005-10-05T23:59:59.000Z
We introduce the neural network approach to global fits of parton distrubution functions. First we review previous work on unbiased parametrizations of deep-inelastic structure functions with faithful estimation of their uncertainties, and then we summarize the current status of neural network parton distribution fits.
Large margin classification in infinite neural networks
Saul, Lawrence K.
Large margin classification in infinite neural networks Youngmin Cho and Lawrence K. Saul, CA 92093-0404 Abstract We introduce a new family of positive-definite kernels for large margin classi- fication in support vector machines (SVMs). These kernels mimic the computation in large neural networks
Neural network approach to parton distributions fitting
Andrea Piccione; Joan Rojo; for the NNPDF Collaboration
2005-10-18T23:59:59.000Z
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.
Tea classification based on artificial olfaction using bionic olfactory neural network
Yang, X L; Fu, J; Lou, Z G; Wang, L Y; Li, G; Freeman, Walter J III
2006-01-01T23:59:59.000Z
conventional artificial neural network (ANN), chaos shoulda con- ventional artificial neural network, BP network, isconventional artificial neural network, it is an accurate
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)
2011-03-15T23:59:59.000Z
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.
Fundamental building blocks for a compact optoelectronic neural network processor
Ruedlinger, Benjamin Franklin, 1976-
2003-01-01T23:59:59.000Z
The focus of this thesis is interconnects within the Compact Optoelectronic Neural Network Processor. The goal of the Compact Optoelectronic Neural Network Processor Project (CONNPP) is to build a small, rugged neural ...
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)
2008-06-12T23:59:59.000Z
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.
Reviews of computing technology: An overview of neural networks
Rainsford, A.E.
1992-02-15T23:59:59.000Z
This report discusses the historical background, models, computer hardware, and uses of neural networks. (LSP)
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
Neural network modelling of hot deformation of austenite
Cambridge, University of
Neural network modelling of hot deformation of austenite Mathew Peet Wolfson College University Neural Networks Neural networks to predict constitutive behaviour 6 6 7 17 20 Experimental Detail 21. Linear regression techniques are not capable of representing the data, however neural networks
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
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
Tampa Electric Neural Network Sootblowing
Mark A. Rhode
2004-09-30T23:59:59.000Z
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.
Aircraft System Identification Using Artificial Neural Networks
Valasek, John
Aircraft System Identification Using Artificial Neural Networks Kenton Kirkpatrick , Jim May Jr linear system identification for aircraft using artificial neural net- works. The output of a linear the correct model. In this paper, a new method of system identification is proposed that uses artificial
Lee, S H; Dornfeld, D A
2007-01-01T23:59:59.000Z
Appli- cation of artificial neural network in laser weldingwith minimal heights. Artificial neural network and non-milling using an artificial neural network with optimized
Exploring Fractional Order Calculus as an Artificial Neural Network Augmentation Samuel Alan Gardner
Dyer, Bill
Exploring Fractional Order Calculus as an Artificial Neural Network Augmentation by Samuel Alan....................................................................................... 4 Artificial Neural Networks DESCRIPTION......................................................................... 22 Neural Network
Yang, Qingsheng; Mwenda, Kevin M; Ge, Miao
2013-01-01T23:59:59.000Z
reasoning and artificial neural network techniques forfactors with artificial neural networks to predict referencefactors with artificial neural networks to predict reference
Paudel, Subodh; Elmtiri, Mohamed; Kling, Wil L; Corre, Olivier Le; Lacarriere, Bruno
2014-01-01T23:59:59.000Z
simulation and artificial neural network for forecastingloads using artificial neural networks, 2001 World Congress,consumption by using artificial neural network, Advances in
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
Imbibition well stimulation via neural network design
Weiss, William (Socorro, NM)
2007-08-14T23:59:59.000Z
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.
Using neural networks to locate pitch accents.
Taylor, Paul A
1995-01-01T23:59:59.000Z
This paper descirbes a technique for finding intonatioanl events, (pitch accents and boundary tones) from waveforms. The technique works in a bottom-up manner by using a recurrent neural network to perform a classification of ...
Parametrizing Compton form factors with neural networks
Kresimir Kumericki; Dieter Mueller; Andreas Schafer
2011-12-08T23:59:59.000Z
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.
CLASSIFICATION OF EXPRESSION PATTERNS USING ARTIFICIAL NEURAL NETWORKS
Ringnér, Markus
Chapter 11 CLASSIFICATION OF EXPRESSION PATTERNS USING ARTIFICIAL NEURAL NETWORKS Markus Ringn´er1 Artificial neural networks in the form of feed-forward networks (ANNs) have emerged as a practical technology
Conceptualization and image understanding by neural networks
Gudipalley, Chandu
1993-01-01T23:59:59.000Z
as that of the associations between concepts into the knowledge base. However a neural network based image understanding system would only need to have conceptual relationships in the knowledge hase. The attributes pertaining to properties are already coded...CONCEPTUALIZATION AND IMAGE UNDERSTANDING BY NEURAL NETWORKS A Thesis by GUDIPALLEY CHANDU Submitted to the Office of Graduate Studies of Texas AkM University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE...
Automatic well log correlation using neural networks
Habiballah, Walid Abdulrahim
1991-01-01T23:59:59.000Z
AUTOMATIC WELL LOG CORRELATION USING NEURAL NETWORKS A Thesis by WALID ABDULHAHIM HABIBALLAH Submitted to the Office of Graduate Studies of Texas AaM University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE... August 1991 Major Subject; Petroleum Engineering AUTOMATIC WELL LOG CORRELATION USING NEURAL NETWORKS A Thesis by WALID ABDULRAHIM HABIBALLAH Approved as to style and content by: R. A. St tzman (Chair of Committee) S. W. Poston (Member) R. R...
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
Dawson, Michael
1368 IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 20, NO. 8, AUGUST 2009 Simple Artificial Neural phenomenon called probability matching (Vulkan 2000). Here, we show that a simple artificial neural network a simple artificial neural network can vary response strengths in accordance with such probability matching
Design and Development of an Artificial Neural Network for Estimation of Formation Permeability
Mohaghegh, Shahab
permeability using artificial neural networks. Neural nets are analog, inherently parallel, distributive
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
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 a trained artificial neural network loses its accuracy when the network is trained again on a different
Parametric Optimization of Artificial Neural Networks for Signal Approximation Applications
Parametric Optimization of Artificial Neural Networks for Signal Approximation Applications J. Lane.thames@gatech.edu randal.abler@gatech.edu dirk.schaefer@me.gatech.edu ABSTRACT Artificial neural networks are used to solve set of configuration parameters for artificial neural networks such that the network's approximation
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)
2004-12-01T23:59:59.000Z
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.
Use of autoassociative neural networks for sensor diagnostics
Najafi, Massieh
2005-02-17T23:59:59.000Z
The new approach for sensor diagnostics is presented. The approach, Enhanced Autoassociative Neural Networks (E-AANN), adds enhancement to Autoassociative Neural Networks (AANN) developed by Kramer in 1992. This enhancement allows AANN to identify...
Modeling of a continuous food process with neural networks
Bullock, David Cole
1995-01-01T23:59:59.000Z
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 ...
Modeling of a continuous food process with neural networks
Bullock, David Cole
1995-01-01T23:59:59.000Z
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 (FFN), a radial basis...
Use of autoassociative neural networks for sensor diagnostics
Najafi, Massieh
2005-02-17T23:59:59.000Z
The new approach for sensor diagnostics is presented. The approach, Enhanced Autoassociative Neural Networks (E-AANN), adds enhancement to Autoassociative Neural Networks (AANN) developed by Kramer in 1992. This enhancement allows AANN to identify...
An introduction to artificial neural networks
C. A. L. Bailer-Jones; R. Gupta; H. P. Singh
2001-02-13T23:59:59.000Z
Artificial neural networks are algorithms which have been developed to tackle a range of computational problems. These range from modelling brain function to making predictions of time-dependent phenomena to solving hard (NP-complete) problems. In this introduction we describe a single, yet very important, type of network known as a feedforward network. This network is a mathematical model which can be trained to learn an arbitrarily complex relationship between a data and a parameter domain, so can be used to solve interpolation and classification problems. We discuss the structure, training and interpretation of these networks, and their implementation, taking the classification of stellar spectra as an example.
Pan Danguang; Gao Yanhua; Song Junlei [School of Civil and Environmental Engineering, University of Science and Technology Beijing, Beijing, 100083 (China)
2010-05-21T23:59:59.000Z
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.
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 a safety lifecycle for artificial neural networks. The lifecycle fo- cuses on managing behaviour
Combinatorial Optimization with Feedback Artificial Neural Networks \\Lambda
Peterson, Carsten
Combinatorial Optimization with Feedback Artificial Neural Networks \\Lambda Carsten Peterson@thep.lu.se Abstract A brief review is given for using feedback artificial neural networks (ANN) to obtain good Neural Networks, OcÂ tober 1995, Paris, France , eds. F. FogelmanÂSoulie and P. Gallinari, EC2 & Cie
Novel Artificial Neural Networks For Remote-Sensing Data Classification
Michel, Howard E.
Novel Artificial Neural Networks For Remote-Sensing Data Classification Xiaoli Tao* and Howard E artificial neural network architectures applied to multi-class classification problems of remote-sensing data. These approaches are 1) a spiking-neural-network model for the partitioning of data into clusters, and 2) a neuron
Using Artificial Neural Networks to Play Pong Luis E. Ramirez
Meeden, Lisa A.
Using Artificial Neural Networks to Play Pong Luis E. Ramirez May 9th, 2014 Abstract This paper examines the possibility of using Artificial Neural Networks to control AI for simple computer games Stanley that evolves artificial neural network topologies simultane- ously with the edge weights[3
Devices and Circuits for Nanoelectronic Implementation of Artificial Neural Networks
Devices and Circuits for Nanoelectronic Implementation of Artificial Neural Networks A Dissertation Implementation of Artificial Neural Networks by ¨Ozg¨ur T¨urel Doctor of Philosophy in Physics and Astronomy. Unfortunately, most artificial neural networks, either software or hardware, do not provide either the speed
Artificial Neural Networks for Recognition of Electrocardiographic Lead Reversal
Peterson, Carsten
Artificial Neural Networks for Recognition of Electrocardiographic Lead Reversal Bo Heden, Iv, which are rule-based,is a diffi- cult task, even for the expert. Artificial neural networks (ANNs) have lack of prop- 100%). The neural networks performed better *an `the er treatment. The pur r se
REGULARIZATION OF A PROGRAMMED RECURRENT ARTIFICIAL NEURAL NETWORK
Meade, Andrew J.
REGULARIZATION OF A PROGRAMMED RECURRENT ARTIFICIAL NEURAL NETWORK Andrew J. Meade, Jr. Department ARTIFICIAL NEURAL NETWORK Andrew J. Meade, Jr. Department of Mechanical Engineering and Materials Science into an artificial neural network architecture. GTR provides a rational means of combining theoretical models
A NOVEL MAP PROJECTION USING AN ARTIFICIAL NEURAL NETWORK
Skupin, André
A NOVEL MAP PROJECTION USING AN ARTIFICIAL NEURAL NETWORK André Skupin Department of Geography is an unsupervised, artificial neural network (ANN) technique. While it should not be mistaken as imitating neural network approach in two ways. First, the paper presents a cartographically informed method
Chaotic time series prediction using artificial neural networks
Bartlett, E.B.
1991-12-31T23:59:59.000Z
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.
1991-01-01T23:59:59.000Z
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.
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
Joy, Mike
An accurate comparison of methods for quantifying variable importance in artificial neural networks, Joy and Death: Assessing variable contributions in neural networks 2 Abstract Artificial neural elements called artificial neural networks (ANNs). Although ANNs were initially developed to better
Dumidu Wijayasekara; Milos Manic; Piyush Sabharwall; Vivek Utgikar
2011-07-01T23:59:59.000Z
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.
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)
2011-10-15T23:59:59.000Z
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.
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
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)
2014-05-15T23:59:59.000Z
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.
Adaptive Neural Networks for Automatic Negotiation
Sakas, D. P.; Vlachos, D. S.; Simos, T. E. [University of Peloponnese, 22100 Tripoli (Greece)
2007-12-26T23:59:59.000Z
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.
Applications of artificial neural networks predicting macroinvertebrates in freshwaters
Lek, Sovan
Applications of artificial neural networks predicting macroinvertebrates in freshwaters Peter L. M Artificial neural networks (ANNs) are non-linear mapping structures that can be applied for predictive P. L suitability models can be very valuable. Data driven methods such as artificial neural net- works (ANNs
Automated Interpretation of Myocardial SPECT Perfusion Images Using Artificial Neural Networks
Peterson, Carsten
Automated Interpretation of Myocardial SPECT Perfusion Images Using Artificial Neural Networks Dan as inputs to multilayer perceptron artificial neural networks. The networks were trained to detect coronary% and not statistically significant. Conclusions: Artificial neural networks can detect coronary artery disease
Davis, Bron
2011-01-01T23:59:59.000Z
algorithm, and Artificial Neural Network." Building andOrtega. "New artificial neural network prediction method fora feedback artificial neural network." Energy and Buildings
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
Conceptualization and image understanding by neural networks
Gudipalley, Chandu
1993-01-01T23:59:59.000Z
. 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)
1997-01-01T23:59:59.000Z
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.
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)
2007-06-13T23:59:59.000Z
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})
Artificial neural network cardiopulmonary modeling and diagnosis
Kangas, L.J.; Keller, P.E.
1997-10-28T23:59:59.000Z
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)
1992-12-31T23:59:59.000Z
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))
1992-01-01T23:59:59.000Z
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
The neural network approach to parton distribution functions
Joan Rojo
2006-07-11T23:59:59.000Z
We introduce the neural network approach to the parametrization of parton distributions. After a general introduction, we present in detail our approach to parametrize experimental data, based on a combination of Monte Carlo methods and neural networks. We apply this strategy first in three different cases: the proton structure function, hadronic tau decays and B meson decay spectra. Finally we describe the neural network approach applied to the parametrization of parton distribution functions, and present results on the nonsinglet parton distribution.
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)
1995-01-01T23:59:59.000Z
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.
1995-12-26T23:59:59.000Z
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.
Applications of Artificial Neural Networks (ANNs) to Rotating Equipment
Sainudiin, Raazesh
, engines), driven equipment (compressors, pumps, mixers, fans, extruders), transmission devices (gears diagnosis, trouble shooting, maintenance, sensor validation, and control. Artificial Neural Network (ANN
Neural networks and their application to nuclear power plant diagnosis
Reifman, J. [Argonne National Lab., IL (United States). Reactor Analysis Div.
1997-10-01T23:59:59.000Z
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.
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 Systems R. E. Bourguet, P. J. Antsaklis, "Artificial Neural Networks in Electric Power Industry. Bourguet, P. J. Antsaklis, "Artificial Neural Networks in Electric Power Industry," Technical Report
A COMPARISON OF ARTIFICIAL NEURAL NETWORKS AND CLUSTER ANALYSIS FOR TYPING BIOMETRICS
A COMPARISON OF ARTIFICIAL NEURAL NETWORKS AND CLUSTER ANALYSIS FOR TYPING BIOMETRICS biometrics, artificial neural networks, cluster analysis, Multi Layer Perceptrons, K- means clustering are clustering techniques and Artificial Neural Networks , in conjunction with data processing to improve
Recognizing targets from infrared intensity scan patterns using artificial neural networks
Barshan, Billur
Recognizing targets from infrared intensity scan patterns using artificial neural networks Tayfun complicating the localization and recognition process. We employ artificial neural networks to deter- mine differentiation; artificial neural networks; optimal brain surgeon; pattern recognition. Paper 080450R received
A RECONFIGURABLE COMPUTING ARCHITECTURE FOR IMPLEMENTING ARTIFICIAL NEURAL NETWORKS ON FPGA
Areibi, Shawki M
A RECONFIGURABLE COMPUTING ARCHITECTURE FOR IMPLEMENTING ARTIFICIAL NEURAL NETWORKS ON FPGA Professor Shawki Areibi Artificial Neural Networks (ANNs), and the backpropagation algorithm in particular NEURAL NETWORKS ON FPGA Kristian Nichols University of Guelph, 2003 Advisor: Professor Medhat Moussa
Wiggins, Vince L.
1996-01-01T23:59:59.000Z
This research evaluates the potential for applying neural networks to Air Force personnel analysis through a review of relevant literature and empirical testing of neural networks in the domain of personnel research. Neural network technology has...
Wiggins, Vince L.
1996-01-01T23:59:59.000Z
This research evaluates the potential for applying neural networks to Air Force personnel analysis through a review of relevant literature and empirical testing of neural networks in the domain of personnel research. Neural network technology has...
Decision-Oriented Environmental Mapping with Radial Basis Function Neural Networks
Gilardi, Nicolas
Decision-Oriented Environmental Mapping with Radial Basis Function Neural Networks V. Demyanov (1 functions neural networks (RBFNN) for spatial predictions. Geostatistical tools for spatial correlation. Artificial neural networks (ANN) offers an alternative way of dealing with spatial information when
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
2010-01-01T23:59:59.000Z
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)
2006-09-15T23:59:59.000Z
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 networks in models of specialization, guild evolution and sympatric speciation
Getz, Wayne M.
Artificial neural networks in models of specialization, guild evolution and sympatric speciation artificial neural networks (ANN) as models for the host recognition system in exploiters, illustrate how
Artificial Neural Network Circuit for Spectral Pattern Recognition
Rasheed, Farah
2013-09-04T23:59:59.000Z
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...
Artificial Neural Network Circuit for Spectral Pattern Recognition
Rasheed, Farah
2013-09-04T23:59:59.000Z
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...
Paraphrastic recurrent neural network language models
Liu, X.; Chen, X.; Gales, M. J. F.; Woodland, P. C.
2015-04-22T23:59:59.000Z
segmentation transducer, T?:?? 1In common with other paraphrase induction methods [1, 18], this scheme can also produce phrase pairs that are non-paraphrastic, for exam- ple, antonyms. However, this is of less concern for language modelling, for which improving... models,” Computer Speech & Language, vol. 21, no. 3, pp. 492–518, 2007. [28] Y. Si, Q. Zhang, T. Li, J. Pan, and Y. Yan (2013), “Prefix tree based n-best list re-scoring for recurrent neural network language model used in speech recognition system...
Applications of Neural Networks in Hadron Physics
Krzysztof M. Graczyk; Cezary Juszczak
2014-09-18T23:59:59.000Z
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.
Electric Power System Anomaly Detection Using Neural Networks
Tronci, Enrico
Electric 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 of an Electric Power System. In this paper, a neural network based approach for novelty detection is presented
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
Shawn Snider; Yuan Qu; Carlos Allende Prieto; Ted von Hippel; Timothy C. Beers; Chistopher Sneden; David L. Lambert
1999-12-19T23:59:59.000Z
New generation large-aperture telescopes, multi-object spectrographs, and large format detectors are making it possible to acquire very large samples of stellar spectra rapidly. In this context, traditional star-by-star spectroscopic analysis are no longer practical. New tools are required that are capable of extracting quickly and with reasonable accuracy important basic stellar parameters coded in the spectra. Recent analyses of Artificial Neural Networks (ANNs) applied to the classification of astronomical spectra have demonstrated the ability of this concept to derive estimates of temperature and luminosity. We have adapted the back-propagation ANN technique developed by von Hippel et al. (1994) to predict effective temperatures, gravities and overall metallicities from spectra with resolving power ~ 2000 and low signal-to-noise ratio. We show that ANN techniques are very effective in executing a three-parameter (Teff,log g,[Fe/H]) stellar classification. The preliminary results show that the technique is even capable of identifying outliers from the training sample.
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
West, Stuart
Y Keywords: adaptation; artificial neural networks; evolutionary constraints; parasitoid; sex ratio by modelling information acquisition and processing using artificial neural networks (ANNs) evolving accordingConstraints on adaptation: explaining deviation from optimal sex ratio using artificial neural
Spikes Synchronization in Neural Networks with Synaptic Plasticity
Borges, Rafael R; Batista, Antonio M; Caldas, Iberê L; Borges, Fernando S; Lameu, Ewandson L
2015-01-01T23:59:59.000Z
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.
Predicting Turbulence Using Partial Least Squares Regression and an Artificial Neural Network
Lakshmanan, Valliappa
Predicting Turbulence Using Partial Least Squares Regression and an Artificial Neural Network #12;Neural Network Neural Network Architecture 6 inputs (the 6 transformed components) 1 output (0 Lakshmanan et. al (OU/NSSL) PLS and NN 8th Conf. on AI, Atlanta, GA 9 / 15 #12;Neural Network Validation
Neural Networks, Radial Basis Functions, and Complexity Mark A. Kon1
Kon, Mark
1 Neural Networks, Radial Basis Functions, and Complexity Mark A. Kon1 Boston University for the non-expert to the theory of artificial neural networks as embodied in current versions of feedforward neural networks. There is a lot of neural network theory which is not mentioned here, including the large
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
Neural Networks, Radial Basis Functions, and Complexity Mark A. Kon 1
Kon, Mark
1 Neural Networks, Radial Basis Functions, and Complexity Mark A. Kon 1 Boston University for the nonexpert to the theory of artificial neural networks as embodied in current versions of feedforward neural networks. There is a lot of neural network theory which is not mentioned here, including the large
Constraint methods for neural networks and computer graphics
Platt, J.C.
1989-01-01T23:59:59.000Z
Both computer graphics and neural networks are related, in that they model natural phenomena. Physically-based models are used by computer graphics researchers to create realistic, natural animation, and neural models are used by neural network researchers to create new algorithms or new circuits. To exploit successfully these graphical and neural models, engineers want models that fulfill designer-specified goals. These goals are converted into mathematical constraints. This thesis presents constraint methods for computer graphics and neural networks. The mathematical constraint methods modify the differential equations that govern the neural or physically-based models. The constraint methods gradually enforce the constraints exactly. This thesis also described application of constrained models to real problems. The first half of this theses discusses constrained neural networks. The desired models and goals are often converted into constrained optimization problems. These optimization problems are solved using first-order differential equations. The applications of constrained neural networks include the creation of constrained circuits, error-correcting codes, symmetric edge detection for computer vision, and heuristics for the traveling salesman problem. The second half of this thesis discusses constrained computer graphics models. In computer graphics, the desired models and goals become constrained mechanical systems, which are typically simulated with second-order differential equations. The Penalty Method adds springs to the mechanical system to penalize violations of the constraints. Rate Controlled Constraints add forces and impulses to the mechanical system to fulfill the constraints with critically damped motion.
Beneficial role of noise in artificial neural networks
Monterola, Christopher [National Institute of Physics, University of the Philippines 1101 Diliman Quezon City (Philippines); Max-Planck Institut fuer Physik Komplexer Systeme Noethnitzerstrasse 38, 01187, Dresden (Germany); Saloma, Caesar [National Institute of Physics, University of the Philippines 1101 Diliman Quezon City (Philippines); Zapotocky, Martin [Max-Planck Institut fuer Physik Komplexer Systeme Noethnitzerstrasse 38, 01187, Dresden (Germany)
2008-06-18T23:59:59.000Z
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.
Estimating photometric redshifts with artificial neural networks
Andrew E. Firth; Ofer Lahav; Rachel S. Somerville
2002-10-21T23:59:59.000Z
A new approach to estimating photometric redshifts - using Artificial Neural Networks (ANNs) - is investigated. Unlike the standard template-fitting photometric redshift technique, a large spectroscopically-identified training set is required but, where one is available, ANNs produce photometric redshift accuracies at least as good as and often better than the template-fitting method. The Bayesian priors on the underlying redshift distribution are automatically taken into account. Furthermore, inputs other than galaxy colours - such as morphology, angular size and surface brightness - may be easily incorporated, and their utility assessed. Different ANN architectures are tested on a semi-analytic model galaxy catalogue and the results are compared with the template-fitting method. Finally the method is tested on a sample of ~ 20000 galaxies from the Sloan Digital Sky Survey. The r.m.s. redshift error in the range z < 0.35 is ~ 0.021.
Galaxies, Human Eyes and Artificial Neural Networks
O. Lahav; A. Naim; R. J. Buta; H. G. Corwin; G. de Vaucouleurs; A. Dressler; J. P. Huchra; S. van den Bergh; S. Raychaudhury; L. Sodre Jr.; M. C. Storrie-Lombardi
1994-12-08T23:59:59.000Z
Quantitative morphological classification of galaxies is important for understanding the origin of type frequency and correlations with environment. But galaxy morphological classification is still mainly done visually by dedicated individuals, in the spirit of Hubble's original scheme, and its modifications. The rapid increase in data on galaxy images at low and high redshift calls for re-examination of the classification schemes and for new automatic methods. Here we show results from the first systematic comparison of the dispersion among human experts classifying a uniformly selected sample of over 800 digitised galaxy images. These galaxy images were then classified by six of the authors independently. The human classifications are compared with each other, and with an automatic classification by Artificial Neural Networks (ANN). It is shown that the ANNs can replicate the classification by a human expert to the same degree of agreement as that between two human experts.
Financial Market Modeling with Quantum Neural Networks
Gonçalves, Carlos Pedro
2015-01-01T23:59:59.000Z
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)
2012-04-01T23:59:59.000Z
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.
Neural Network Approach to Locating Cryptography in Object Code
Jason L. Wright; Milos Manic
2009-09-01T23:59:59.000Z
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.
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
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
Evolution of Memory in Reactive Artificial Neural Networks
Chung, Ji Ryang
2012-07-16T23:59:59.000Z
in the context of evolution: how reactive agents could have evolved into cognitive ones with internalized memory? This study strives to find an answer to the question by simulating neuroevolution on artificial neural networks, with the hypothesis...
Digital neural network-based modeling technique for extrusion processes
Jang, Won-Hyouk
2001-01-01T23:59:59.000Z
and market conditions. In order to develop reliable and well-performing advanced process monitoring and diagnostic systems for achieving improved product quality and cost-effective operation, the neural network-based modeling technique for the extrusion...
A neural network approach to snack quality evaluation
Sayeed, Mohammad Shaheen
1994-01-01T23:59:59.000Z
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...
Structural Impairment Detection Using Arrays of Competitive Artificial Neural Networks
Story, Brett
2012-07-16T23:59:59.000Z
algorithm development, and the integration of data acquisition and impairment detection tools. Ultimately, data streams from the Salmon Bay Bridge are autonomously recorded and interrogated by competitive arrays of artificial neural networks for patterns...
Neural network calibration for miniature multi-hole pressure probes
Vijayagopal, Rajesh
1998-01-01T23:59:59.000Z
A robust and accurate neural network based algorithm phics. for the calibration of miniature multi-hole pressure probes has been developed and a detailed description of its features and use is presented. The code that was developed was intended...
Neural Network Parameterizations of Electromagnetic Nucleon Form Factors
Krzysztof M. Graczyk; Piotr Plonski; Robert Sulej
2010-08-25T23:59:59.000Z
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)
2009-05-23T23:59:59.000Z
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
2011-01-01T23:59:59.000Z
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 ...
Evolution of Memory in Reactive Artificial Neural Networks
Chung, Ji Ryang
2012-07-16T23:59:59.000Z
in the context of evolution: how reactive agents could have evolved into cognitive ones with internalized memory? This study strives to find an answer to the question by simulating neuroevolution on artificial neural networks, with the hypothesis...
Hybrid digital signal processing and neural networks applications in PWRs
Eryurek, E.; Upadhyaya, B.R.; Kavaklioglu, K.
1991-01-01T23:59:59.000Z
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.
1991-12-31T23:59:59.000Z
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.
A portable neural network approach to vehicle tracking
Miller, Kelly Maxwell
1994-01-01T23:59:59.000Z
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...
Neural network based design of cellular manufacturing systems
Ramachandran, Satheesh
1990-01-01T23:59:59.000Z
NEURAL NETWORK BASED DESIGN OF CELLULAR MANUFACTURING SYSTEMS A Thesis by SATHEESH RAMACHANDRAN Submitted to the Office of Graduate Studies of' Texas ASM University in partial fulfilltnent of the requirements for the degree of MASTER... OF SCIENCE December 1990 Major Subject: Industrial Engineering NEURAL NETWORK BASED DESIGN OF CELLULAR MANUFACTURING SYSTEMS A Thesis by SATHEESH RAMACHAiVDRAN Approved as to style and content by: Ce r O. Malav (Chair of Committee) T. Hsing...
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)
2008-11-06T23:59:59.000Z
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)
2009-03-03T23:59:59.000Z
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.
Paris-Sud XI, Université de
Artificial Neural Network Technology: for Classification and Cartography of Scientific Artificial Neural Networks (ANNs) to extend NEURODOC into a neural platform for the cluster analysis with the aid of neural networks (models which are essentially non-linear and threshold-driven). In the design
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
2011-01-01T23:59:59.000Z
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 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)
2008-06-12T23:59:59.000Z
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.
Neural Network Based Intelligent Sootblowing System
Mark Rhode
2005-04-01T23:59:59.000Z
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.
Fast cosmological parameter estimation using neural networks
T. Auld; M. Bridges; M. P. Hobson; S. F. Gull
2007-09-17T23:59:59.000Z
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.
Spagnolo, Filippo
431 On Teaching Quality Improvement of a Mathematical Topic Using Artificial Neural Networks inspired by simulation by Artificial Neural Networks (ANNs) applied recently for evaluation of phonics performance. Keywords: Artificial Neural Networks, Learning Performance Evaluation, Computer Aided Learning
A HARDWARE/SOFTWARE CO-DESIGN APPROACH FOR FACE RECOGNITION BY ARTIFICIAL NEURAL NETWORKS
Areibi, Shawki M
#12;A HARDWARE/SOFTWARE CO-DESIGN APPROACH FOR FACE RECOGNITION BY ARTIFICIAL NEURAL NETWORKS NETWORKS Xiaoguang Li University of Guelph, 2004 Advisor: Dr. Shawki M. Areibi Artificial Neural Networks Artificial Neural Networks on FPGAs. This work provided a very good beginning to my research. Special thanks
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
Protein Sequence Classification Using Probabilistic Motifs and Neural Networks
Blekas, Konstantinos
Protein Sequence Classification Using Probabilistic Motifs and Neural Networks Konstantinos Blekas classification is the sequence encoding scheme that must be used in order to feed the network. To deal with this prob- lem we propose a method that maps a protein sequence into a numerical feature space using
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)
2011-05-25T23:59:59.000Z
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.
Strategies for Spectral Profile Inversion using Artificial Neural Networks
H. Socas-Navarro
2004-10-23T23:59:59.000Z
This paper explores three different strategies for the inversion of spectral lines (and their Stokes profiles) using artificial neural networks. It is shown that a straightforward approach in which the network is trained with synthetic spectra from a simplified model leads to considerable errors in the inversion of real observations. This problem can be overcome in at least two different ways that are studied here in detail. The first method makes use of an additional pre-processing auto-associative neural network to project the observed profile into the theoretical model subspace. The second method considers a suitable regularization of the neural network used for the inversion. These new techniques are shown to be robust and reliable when applied to the inversion of both synthetic and observed data, with errors typically below $\\sim$100 G.
Neural networks using two-component Bose-Einstein condensates
Tim Byrnes; Shinsuke Koyama; Kai Yan; Yoshihisa Yamamoto
2012-11-16T23:59:59.000Z
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.
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
Predicting Turbulence using Partial Least Squares Regression and an Artificial Neural Network
Lakshmanan, Valliappa
Predicting Turbulence using Partial Least Squares Regression and an Artificial Neural Network in the dataset. Then, the transformed data are pre- sented to a neural network whose output node has a sigmoid
Use of neural networks to correlate enzymatic hydrolysis with biomass properties
Narayan, Ramasubramanian
2001-01-01T23:59:59.000Z
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 previously developed empirical...
Reservoir characterization using seismic attributes, well data, and artificial neural networks
Toinet, Sylvain
2001-01-01T23:59:59.000Z
This study reports an investigation of the potentialities of artificial neural networks in the field of reservoir characterization. A first step has been the review of theoretical principles involved in neural networks computations, in order...
Representing liquid-vapor equilibria of Ternary systems using neural networks
Swisher, Mathew M
2015-01-01T23:59:59.000Z
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
Storage capacity and retrieval time of small-world neural networks
Oshima, Hiraku; Odagaki, Takashi [Department of Physics, Kyushu University, Fukuoka 812-8581 (Japan)
2007-09-15T23:59:59.000Z
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.
Real-time neural network earthquake profile predictor
Leach, Richard R. (Castro Valley, CA); Dowla, Farid U. (Castro Valley, CA)
1996-01-01T23:59:59.000Z
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.
1996-02-06T23:59:59.000Z
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.
Using a neural network for abnormal event identification in BWRs
Ohga, Yukiharu; Seki, Hiroshi (Hitachi Ltd., Ibaraki (Japan))
1991-01-01T23:59:59.000Z
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.
Paris-Sud XI, Université de
on artificial neural networks, neural classifier and weather predictions. Application in an urban atmosphere - will be solved. Keywords: Artificial neural network; Multilayer Perceptron; ozone modelling; statistical stepwise and Software 22, 9 (2007) 1261-1269" DOI : 10.1016/j.envsoft.2006.08.002 #12;Abstract A neural network combined
Nuclear power plant fault-diagnosis using artificial neural networks
Kim, Keehoon; Aljundi, T.L.; Bartlett, E.B.
1992-01-01T23:59:59.000Z
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.
Paris-Sud XI, Université de
1 Optimization of an artificial neural network dedicated to the multivariate forecasting of daily Ajaccio, France Abstract. This paper presents an application of Artificial Neural Networks (ANNs Artificial Neural Networks (ANNs) which are a popular artificial intelligence technique in the forecasting
Fernandez, Thomas
Comparative application of artificial neural networks and genetic algorithms for multivariate time of artificial neural networks and genetic algorithms in terms of forecasting and understanding of algal blooms-a, Microcystis, short-term prediction, artificial neural network model, genetic algorithm model, rule sets
Computational subunits of visual cortical neurons revealed by artificial neural networks
Lau, Brian
Computational subunits of visual cortical neurons revealed by artificial neural networks Brian Lau neurons of the cat to spatiotemporal random-bar stimuli and trained artificial neural networks to predict with a simple functional model for complex cells and demonstrate the usefulness of the neural network method
Application of Artificial Neural Network in Social Computing in the Context of Third World Countries
Ghosh, Joydeep
Application of Artificial Neural Network in Social Computing in the Context of Third World - In the last decade, applications associated with artificial neural network (ANN) has been gaining popularity amount of data, application of computational tools like Artificial Neural Network, Association rules
Evolving artificial neural networks to control chaotic systems Eric R. Weeks* and John M. Burgess
Weeks, Eric R.
Evolving artificial neural networks to control chaotic systems Eric R. Weeks* and John M. Burgess 78712 Received 7 April 1997 We develop a genetic algorithm that produces neural network feedback of unstable fixed points. This is the first dimension- independent algorithm that produces neural network
MODELING PHYTOPLANKTON ABUNDANCE IN SAGINAW BAY, LAKE HURON: USING ARTIFICIAL NEURAL NETWORKS; phytoplankton Abbreviations: ANN, artificial neural network; ClÀ , chloride; DOC, dissolved organic carbon; Kd Phytoplankton abundance, as chl a, in Saginaw Bay, Lake Huron was modeled using arti- ficial neural networks
May 24, 2012 18:36 manuscript Data Processing using Artificial Neural Networks
Paris-Sud XI, Université de
May 24, 2012 18:36 manuscript Data Processing using Artificial Neural Networks to Improve uncertainties introduced by the 4D-CT scan, we conveniently treated data using artificial neural networks. More step are then used to build a training set for another artificial neural network that learns the lung
Smith, Alice E.
A Predictive Model for Slip Resistance Using Artificial Neural Networks Janet M. Twomey, IIE Artificial Neural Networks Why This Paper is Important Slips and falls are a serious ergonomic problem a slip resistance testing device were used to develop an artificial neural network model which predicts
Vermont, University of
Stream classification using hierarchical artificial neural networks: A fluvial hazard management-in-Chief Keywords: Stream classification Artificial neural networks Kohonen self-organizing maps Counterpropagation. In this research, we apply non-parametric, clus- tering and classification artificial neural networks to assimilate
Artificial neural network discrimination of black-capped chickadee (Poecile atricapillus) call notes
Dawson, Michael
Artificial neural network discrimination of black-capped chickadee (Poecile atricapillus) call; revised 11 April 2006; accepted 12 May 2006 Artificial neural networks were trained to discriminate number s : 43.80Ka., 43.80.Lb, 43.80.Jz JAS Pages: 11111117 I. INTRODUCTION An artificial neural network
Automated Interpretation of Myocardial SPECT Perfusion Images Using Artificial Neural Networks
Peterson, Carsten
Automated Interpretation of Myocardial SPECT Perfusion Images Using Artificial Neural Networks Dan. Conclusion: Artificial neural networks can detect CAD in myocardial bull's-eye scintigrams with such a high significant potential. Key Words: diagnosis; computer-assisted; artificial intelligence; neural networks
Advances in ungauged streamflow prediction using artificial neural networks Lance E. Besaw a
Vermont, University of
Advances in ungauged streamflow prediction using artificial neural networks Lance E. Besaw-John Chang, Associate Editor Keywords: Ungauged streamflow prediction Artificial neural networks Time and test two artificial neural networks (ANNs) to forecast streamflow in unga- uged basins. The model
Paris-Sud XI, Université de
SEQUENTIAL IMPORTANCE SAMPLING BASED ON A COMMITTEE OF ARTIFICIAL NEURAL NETWORKS FOR POSTERIOR) is used in this study to filter the output distribution from a committee of Artificial Neural Networks at discrete time steps. KEYWORDS : model identification, artificial neural network, committee, sequential
Zhang, Wang, and Wei 1 An Artificial Neural Network Method for Length-based
Wang, Yinhai
Zhang, Wang, and Wei 1 An Artificial Neural Network Method for Length-based Vehicle Classification this problem, we develop an artificial neural network method to estimate classified vehicle volumes directly stations on I-5 over a long duration. The results show that the proposed artificial neural network model
Paris-Sud XI, Université de
Transitional Modeling of Building Heating Energy Demand Using Artificial1 Neural Network2 Subodh Paudel a artificial12 neural network. In addition, novel pseudo dynamic transitional model is introduced, which Institution15 building and compared its results with static and other pseudo dynamic neural network models
Dynamic Pattern Recognition in Sport by Means of Artificial Neural Networks
Perl, Jürgen
Dynamic Pattern Recognition in Sport by Means of Artificial Neural Networks Jürgen Perl, Peter to store these data but to transform them into useful information. Artificial Neural Networks turn out the data. This is the point where Artificial Neural Networks can become extremely helpful: They are able
Evolving artificial neural networks to control chaotic systems Eric R. Weeks* and John M. Burgess +
Weeks, Eric R.
Evolving artificial neural networks to control chaotic systems Eric R. Weeks* and John M. Burgess 78712 ~Received 7 April 1997! We develop a genetic algorithm that produces neural network feedback a robust method of train ing neural networks to control chaos. The method makes no assumptions about
Artificial neural networks in models of specialisation, guild evolution and sympatric speciation
Getz, Wayne M.
11 Artificial neural networks in models of specialisation, guild evolution and sympatric speciation 1/12/10 6:05:26pm page 236 Modelling Perception in Artificial Neural Networks, ed. C. R. Tosh and G artificial neural networks as models for the plant recognition mechanism in insects (Holmgren & Getz,
Combining Q-Learning with Artificial Neural Networks in an Adaptive Light Seeking Robot
Meeden, Lisa A.
Combining Q-Learning with Artificial Neural Networks in an Adaptive Light Seeking Robot Steve Dini an artificial neural network as a function approximator and eliminate the need for an explicit table. 1 limitations of using tables as well as the viability of using artificial neural networks to approximate
Multi-point tidal prediction using artificial neural network with tide-generating forces
Multi-point tidal prediction using artificial neural network with tide-generating forces Hsien Available online 23 June 2006 Abstract This paper presents a neural network model of simulating tides Elsevier B.V. All rights reserved. Keywords: Neural networks; Tides; Tide-generating forces; Harmonic
Michel Verleysen Altran 18/11/2002 -1 Artificial Neural Networks
Verleysen, Michel
1 Michel Verleysen Altran 18/11/2002 - 1 Artificial Neural Networks: general overview and specific: Introduction pwhy "artificial neural networks" (ANN) ? pwhat are ANNs useful for? ? plearning generalization Verleysen Altran 18/11/2002 - 3 Content p Part I: Introduction pwhy "artificial neural networks" (ANN
1 Artificial Neural Networks and Hidden Markov Models for Predicting the Protein Structures advice on the development of this project #12;2 Artificial Neural Networks and Hidden Markov Models learning methods: artificial neural networks (ANN) and hidden Markov models (HMM) (Rost 2002; Karplus et al
"Least Squares Fitting" Using Artificial Neural Networks YARON DANON and MARK J. EMBRECHTS
Danon, Yaron
"Least Squares Fitting" Using Artificial Neural Networks YARON DANON and MARK J. EMBRECHTS process changes the internal parameters (weights) of the network such that the neural net can represent a backpropagation fit to various continuous functions will be presented, showing properties of neural network fitted
Automatic classification of Sleep Stages on a EEG signal by Artificial Neural Networks
Paris-Sud XI, Université de
Automatic classification of Sleep Stages on a EEG signal by Artificial Neural Networks Nizar of modeling and design to improve the performances of our tool. Key-Words: - Artificial Neural Networks data analysis tools. Concerning this latter point, we have proposed that Artificial Neural Networks
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
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
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 and Paul B. Watta The Computation and Neural Networks Laboratory Department of Electrical & Computer in the computational capabilities of recurrent neural networks and their relation to finite state machines
Peterson, Carsten
May 1990 LU TP 908 Using Neural Networks to Identify Jets Leif L¨onnblad 1 , Carsten Peterson 2 Lund, Sweden Nuclear Physics B 349, 675 (1991) Abstract: A neural network method for identifying with what is expected from vertex detectors. We also speculate on how the neural network method can be used
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
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))
1989-01-01T23:59:59.000Z
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.
A neural network approach to burn-in
Clifford, Nancy Lynn
1995-01-01T23:59:59.000Z
A NEURAL NETWORK APPROACH TO BURN-IN A Thesis by NANCY LYNN CLIFFORD 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 August 1995 Major... Subject: Industrial Enginccring A NEURAL NETWORK APPROACH TO BURN-IN A Thesis NANCY LYNN CLIFFORD Submitted to Texas A&M University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Approved as to style and content by...
Mathematically Reduced Chemical Reaction Mechanism Using Neural Networks
Nelson Butuk
2004-12-01T23:59:59.000Z
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
2007-08-31T23:59:59.000Z
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.
Morphological Classification of Galaxies Using Artificial Neural Networks
Nicholas M. Ball
2001-10-22T23:59:59.000Z
The results of morphological galaxy classifications performed by humans and by automated methods are compared. In particular, a comparison is made between the eyeball classifications of 454 galaxies in the Sloan Digital Sky Survey (SDSS) commissioning data (Shimasaku et al. 2001) with those of supervised artificial neural network programs constructed using the MATLAB Neural Network Toolbox package. Networks in this package have not previously been used for galaxy classification. It is found that simple neural networks are able to improve on the results of linear classifiers, giving correlation coefficients of the order of 0.8 +/- 0.1, compared with those of around 0.7 +/- 0.1 for linear classifiers. The networks are trained using the resilient backpropagation algorithm, which, to the author's knowledge, has not been specifically used in the galaxy classification literature. The galaxy parameters used and the network architecture are both important, and in particular the galaxy concentration index, a measure of the concentration of light towards the centre of the galaxy, is the most significant parameter. Simple networks are briefly applied to 29,429 galaxies with redshifts from the SDSS Early Data Release. They give an approximate ratio of types E/S0:Sp:Irr of 14 +/- 5 : 86 +/- 12 : 0 +/- 0.1, which broadly agrees with the well known approximate ratios of 20:80:1 observed at low redshift.
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)
2010-07-15T23:59:59.000Z
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
2007-01-16T23:59:59.000Z
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.
Cambridge, University of
of reality. Neural networks form a general method of non{linear regression. Their exibility enables them non{linear function is tted to experimental data, Fig 4.1 as in linear regression, the input variable are derived. The general form of the equation developed using linear regression is a sum of the products
Neural network model of creep strength of austenitic stainless steels
Cambridge, University of
is a parameterised non-linear model which can be used to perform regression, in which case, a very ¯ exible, non-linear of the problems encoun- tered with linear regression. In the present study, neural network analysis was applied with a constant input set to unity. Any non-linear function can be used at the hidden units (as long
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
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
Nonlinear Flight Control Using Neural Networks and Feedback Linearization
Nonlinear Flight Control Using Neural Networks and Feedback Linearization Byoung So0 Kim1, Anthony approaches to aircraft flight control involve linearization of these dynamicsabout a set of pre eliminate many of the undesirable features of linear control. Control of nonlinear systems by inverting
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)
1991-01-01T23:59:59.000Z
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
1991-12-31T23:59:59.000Z
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.
1994-01-01T23:59:59.000Z
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.
NEURAL NETWORK ASSISTED NONLINEAR CONTROLLER FOR A BIOREACTOR
Efe, Mehmet Önder
343 NEURAL NETWORK ASSISTED NONLINEAR CONTROLLER FOR A BIOREACTOR Mehmet Önder Efe Electrical Istanbul, 80815, Turkey ABSTRACT In this paper, design of a nonlinear controller for a Bioreactor Benchmark of such a simplified model is unlikely to result in a satisfactory performance. A prime example is a bioreactor
Practical Variational Inference for Neural Networks Alex Graves
Toronto, University of
are energy-based models such as restricted Boltzmann machines [24] whose log- loss is intractable. 1 #12-loss parametric model--which includes most neural networks1 Variational inference can be reformulated as the optimisation of a Minimum Description length (MDL; [21]) loss function; indeed it was in this form
Optimisation of Neural Network for Charpy Toughness of Steel Welds
Cambridge, University of
Optimisation of Neural Network for Charpy Toughness of Steel Welds Jun Hak, Pak #12;Contents Introduction Modelling of output Bias in Models Interpass temperature effect Results Discussion Conclusions #12) Measured energy absorbed by a standard s ample during fracture Never be negative ! #12;Example of non
Optimisation of Neural Network for Charpy Toughness of Steel Welds
Cambridge, University of
Optimisation of Neural Network for Charpy Toughness of Steel Welds Jun Hak, Pak #12;Contents Introduction Modelling of output Bias in Models Interpass temperature effectInterpass temperature effect unphysical values PossiblePossible Charpy toughness (Impact toughness) Measured energy absorbed by a standard
Neural Network Control of a Pneumatic Robot Ted Hesselroth
Duckett, Tom
Neural Network Control of a Pneumatic Robot Arm Ted Hesselroth , Kakali Sarkar , P. Patrick van der been employed to control a five-joint pneu- matic robot arm and gripper through feedback from two video cameras. The pneumatically driven robot arm (SoftArm) employed in this inves- tigation shares essential
A STATISTICAL NEURAL NETWORK FRAMEWORK FOR RISK MANAGEMENT PROCESS
Basili, Victor R.
both prior and posterior probability. The overall risk evaluation, calculated by those two that the project succeeds? How can we evaluate whether risks actually have impacted on the achievementA STATISTICAL NEURAL NETWORK FRAMEWORK FOR RISK MANAGEMENT PROCESS From the Proposal to its
Language discrimination and clustering via a neural network approach
Mariano, Angelo; Pascazio, Saverio
2015-01-01T23:59:59.000Z
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
2011-10-17T23:59:59.000Z
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.
Wind Power Plant Prediction by Using Neural Networks: Preprint
Liu, Z.; Gao, W.; Wan, Y. H.; Muljadi, E.
2012-08-01T23:59:59.000Z
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
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
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 technology for inflow forecasting, because of (1) their success in power load forecasting 1- 6 , and (2
Transmission Line Boundary Protection Using Wavelet Transform and Neural Network
use data from two ends. The non-unit protection such as distance relay, can not protect the entire end only. In this case, the relay at one end can protect the entire line length with no intentional1 Transmission Line Boundary Protection Using Wavelet Transform and Neural Network Nan Zhang
Hybrid Neural Network for Gas Analysis Measuring System Kazimierz Brudzewski
Osowski, Stanislaw
mixtures of air with these four pollutants. The signals obtained from the sensors have been processed using with hybrid neural network, can be used to determine the individual analyte concentrations in the mixture as pollutants: carbon oxide, methane, propane/buthane and methanol vapour. A small array of five semiconductor
Optimisation of Concentrating Solar Thermal Power Plants with Neural Networks
Ábrahám, Erika
Optimisation of Concentrating Solar Thermal Power Plants with Neural Networks Pascal Richter1, Germany 2 Fraunhofer Institute for Solar Energy Systems, Freiburg, Germany Abstract. The exploitation of solar power for energy supply is of in- creasing importance. While technical development mainly takes
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
Characterization of Shape Memory Alloys Using Artificial Neural Networks
Valasek, John
Characterization of Shape Memory Alloys Using Artificial Neural Networks James V. Henrickson , Kenton Kirkpatrick, and John Valasek Texas A&M University, College Station, Texas 77843-3141 Shape memory of shape memory alloys, however, is often com- plicated by their hysteretic, non-linear, thermo
Secret sharing using artificial neural network
Alkharobi, Talal M.
2004-11-15T23:59:59.000Z
7 Structure of a human brain neural cell ........................................................ 52 8 Structure of an artificial neuron................................................................... 53 9 ANN for 6 shareholders...) ............................ 67 8 Possible set of weights to solve the example with the 2 bits output ANN.. 66 9 Shares given to shareholders in second solution (2-bit output ANN)......... 67 10 Classifying set of shareholders as qualified and unqualified in 3 out...
Predicting stream water quality using artificial neural networks (ANN)
Bowers, J.A.
2000-05-17T23:59:59.000Z
Predicting point and nonpoint source runoff of dissolved and suspended materials into their receiving streams is important to protecting water quality and traditionally has been modeled using deterministic or statistical methods. The purpose of this study was to predict water quality in small streams using an Artificial Neural Network (ANN). The selected input variables were local precipitation, stream flow rates and turbidity for the initial prediction of suspended solids in the stream. A single hidden-layer feedforward neural network using backpropagation learning algorithms was developed with a detailed analysis of model design of those factors affecting successful implementation of the model. All features of a feedforward neural model were investigated including training set creation, number and layers of neurons, neural activation functions, and backpropagation algorithms. Least-squares regression was used to compare model predictions with test data sets. Most of the model configurations offered excellent predictive capabilities. Using either the logistic or the hyperbolic tangent neural activation function did not significantly affect predicted results. This was also true for the two learning algorithms tested, the Levenberg-Marquardt and Polak-Ribiere conjugate-gradient descent methods. The most important step during model development and training was the representative selection of data records for training of the model.
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
Artificial Neural Networks for Solving Ordinary and Partial Differential Equations
I. E. Lagaris; A. Likas; D. I. Fotiadis
1997-05-19T23:59:59.000Z
We present a method to solve initial and boundary value problems using artificial neural networks. A trial solution of the differential equation is written as a sum of two parts. The first part satisfies the boundary (or initial) conditions and contains no adjustable parameters. The second part is constructed so as not to affect the boundary conditions. This part involves a feedforward neural network, containing adjustable parameters (the weights). Hence by construction the boundary conditions are satisfied and the network is trained to satisfy the differential equation. The applicability of this approach ranges from single ODE's, to systems of coupled ODE's and also to PDE's. In this article we illustrate the method by solving a variety of model problems and present comparisons with finite elements for several cases of partial differential equations.
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)
1992-01-01T23:59:59.000Z
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
1992-12-31T23:59:59.000Z
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.
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)
2011-03-24T23:59:59.000Z
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.
Vermont, University of
July 2007; published 8 November 2007. [1] A novel data-driven artificial neural network (ANN and spatial estimation with multiple data types using artificial neural networks, Water Resour. Res., 43, WStochastic simulation and spatial estimation with multiple data types using artificial neural
Oldenburg, Carl von Ossietzky Universität
marine waters with fluorescence spectroscopy and of using artificial neural networks for data interpre with an artificial neural network. The results demonstrate the possibility of estimating oil concentrationsFluorescence diagnostics of oil pollution in coastal marine waters by use of artificial neural
Lek, Sovan
, Greece Abstract Convinced by the predictive quality of artificial neural network (ANN) models in ecology. # 2002 Published by Elsevier Science B.V. Keywords: Artificial neural networks; Backpropagation; NonReview and comparison of methods to study the contribution of variables in artificial neural
Chaubey, Indrajeet
considerable interest in developing methods for uncertainty analysis of artificial neural network (ANN) models and parametric uncertainty in artificial neural network hydrologic models, Water Resour. Res., 43, W10407, doi:10A simplified approach to quantifying predictive and parametric uncertainty in artificial neural
A DISCRETE APPROACH TO CONSTRUCTIVE NEURAL NETWORK
Obradovic, Zoran
and the Paragon supercomputer using p4 are in agreement with analytical speed-up estimates and the architecture is a weighted graph of simple processing units (or neurons). The interconnection graph of a feed-forward network)) involves modi cation of the interconnection weights between neurons on a pre-speci ed network. Determining
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
Liblit, Ben
artificial neural networks as part of a multistrategy learning system, there must be a way for neural; 1 Introduction Artificial neural networks (ANNs) have proven to be a powerful and general technique Knowledge Using Neural Networks Geoffrey G. Towell Jude W. Shavlik University of Wisconsin --- Madison 1210
Boyer, Edmond
such as energy consumption, throughput, delay, coverage, etc. Also many schemes have been proposed in order. In fact, it is sufficient to consider a different input for the neural network to aim to a different monitoring and smart agriculture [1]. One of the fundamental issue in such a network is coverage. It is used
Feasibility of using neural networks as a level 2 calorimeter trigger for jet tagging
Handler, T.; Neis, E. [Tennessee Univ., Knoxville, TN (United States). Dept. of Physics; Glover, C.; Gabriel, T.; Saini, S. [Oak Ridge National Lab., TN (United States)
1992-12-31T23:59:59.000Z
Several of expected decay modes of the Higgs particle will result in jet formation. We propose to incorporate a second level trigger into the SDC detector, using neural network VSLI hardware, to tag such Higgs decay modes. The input to the neural network will be the energy depositions in both the barrel and endcap regions of the calorimeter. The neural network`s output would be a value representing the degree of correlation between the observed energy distribution and the type of physical scattering that has occurred. Preliminary results indicate that neural networks may be of use in tagging jet decays of the Higgs particle.
Adaptive model predictive process control using neural networks
Buescher, Kevin L. (Los Alamos, NM); Baum, Christopher C. (Mazomanie, WI); Jones, Roger D. (Espanola, NM)
1997-01-01T23:59:59.000Z
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.
Measuring photometric redshifts using galaxy images and Deep Neural Networks
Hoyle, Ben
2015-01-01T23:59:59.000Z
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, K.L.; Baum, C.C.; Jones, R.D.
1997-08-19T23:59:59.000Z
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
1990-01-01T23:59:59.000Z
, 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, C.Y.
1996-07-23T23:59:59.000Z
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, Chi Y. (San Francisco, CA)
1996-01-01T23:59:59.000Z
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)
1997-01-01T23:59:59.000Z
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 circuit for forming synapses in neural networks
Fu, C.Y.
1997-02-11T23:59:59.000Z
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.
Re-evaluation of the Gottfried sum using neural networks
Riccardo Abbate; Stefano Forte
2005-11-18T23:59:59.000Z
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
2005-11-11T23:59:59.000Z
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.
Character displacement and the evolution of mate choice: an artificial neural network approach
Ryan, Michael J.
Character displacement and the evolution of mate choice: an artificial neural network approach preferences for aspects of conspecific male signals. We used artificial neural network models to simulate varied in their preferences for aspects of conspecific male signals. When we tested networks
Random Weights Search in CompressedNeural Networks Using OverdeterminedPseudoinverse
Wilamowski, Bogdan Maciej
regions by variable tabu list size. Overvieu, of combination of neural network gradient search and TabuRandom Weights Search in CompressedNeural Networks Using OverdeterminedPseudoinverse Milos Manic is reduction of weight set. Second phase is gradient calculation on such compressed network. Search for weights
Nuclear power plant fault-diagnosis using artificial neural networks
Kim, Keehoon; Aljundi, T.L.; Bartlett, E.B.
1992-12-31T23:59:59.000Z
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)
2008-10-07T23:59:59.000Z
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.
Feasibility of using neural networks as a level 2 calorimeter trigger for jet tagging
Handler, T.; Neis, E. (Tennessee Univ., Knoxville, TN (United States). Dept. of Physics); Glover, C.; Gabriel, T.; Saini, S. (Oak Ridge National Lab., TN (United States))
1992-01-01T23:59:59.000Z
Several of expected decay modes of the Higgs particle will result in jet formation. We propose to incorporate a second level trigger into the SDC detector, using neural network VSLI hardware, to tag such Higgs decay modes. The input to the neural network will be the energy depositions in both the barrel and endcap regions of the calorimeter. The neural network's output would be a value representing the degree of correlation between the observed energy distribution and the type of physical scattering that has occurred. Preliminary results indicate that neural networks may be of use in tagging jet decays of the Higgs particle.
Mathematically Reduced Chemical Reaction Mechanism Using Neural Networks
Nelson Butuk
2006-09-21T23:59:59.000Z
This is an annual technical report for the work done over the last year (period ending 9/30/2005) 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 significant development made in developing a truly meshfree computational fluid dynamics (CFD) flow solver to be coupled to NPCA. First, the procedure of obtaining nearly analytic accurate first order derivatives using the complex step method (CSM) is extended to include computation of accurate meshfree second order derivatives via a theorem described in this report. Next, boosted generalized regression neural network (BGRNN), described in our previous report is combined with CSM and used to obtain complete solution of a hard to solve wave dominated sample second order partial differential equation (PDE): the cubic Schrodinger equation. The resulting algorithm is a significant improvement of the meshfree technique of smooth particle hydrodynamics method (SPH). It is suggested that the demonstrated meshfree technique be termed boosted smooth particle hydrodynamics method (BSPH). Some of the advantages of BSPH over other meshfree methods include; it is of higher order accuracy than SPH; compared to other meshfree methods, it is completely meshfree and does not require any background meshes; It does not involve any construction of shape function with their associated solution of possibly ill conditioned matrix equations; compared to some SPH techniques, no equation for the smoothing parameter is required; finally it is easy to program.
Topological and Dynamical Complexity of Random Neural Networks
Gilles Wainrib; Jonathan Touboul
2013-03-15T23:59:59.000Z
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.
Pattern classification and associative recall by neural networks
Chiueh, Tzi-Dar.
1989-01-01T23:59:59.000Z
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
Paris-Sud XI, Université de
system reliability, artificial neural network, quadratic response surface 1. INTRODUCTION Modern nuclearArtificial Neural Networks and quadratic Response Surfaces for the functional failure analysis of a thermal-hydraulic passive system George Apostolakisa , Nicola Pedronib , Enrico Ziob* a Massachusetts
Neural Network Based Montioring and Control of Fluidized Bed.
Bodruzzaman, M.; Essawy, M.A.
1996-04-01T23:59:59.000Z
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.
Structural Health Monitoring Using Neural Network Based Vibrational System Identification
Sofge, Donald A
2007-01-01T23:59:59.000Z
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.
An artificial neural network application on nuclear charge radii
S. Akkoyun; T. Bayram; S. O. Kara; A. Sinan
2012-12-27T23:59:59.000Z
The artificial neural networks (ANNs) have emerged with successful applications in nuclear physics as well as in many fields of science in recent years. In this paper, by using (ANNs), we have constructed a formula for the nuclear charge radii. Statistical modeling of nuclear charge radii by using ANNs has been seen as to be successful. Also, the charge radii, binding energies and two-neutron separation energies of Sn isotopes have been calculated by implementing of the new formula in Hartree-Fock-Bogoliubov (HFB) calculations. The results of the study shows that the new formula is useful for describing nuclear charge radii.
An artificial neural network application on nuclear charge radii
Akkoyun, S; Kara, S O; Sinan, A
2013-01-01T23:59:59.000Z
The artificial neural networks (ANNs) have emerged with successful applications in nuclear physics as well as in many fields of science in recent years. In this paper, by using (ANNs), we have constructed a formula for the nuclear charge radii. Statistical modeling of nuclear charge radii by using ANNs has been seen as to be successful. Also, the charge radii, binding energies and two-neutron separation energies of Sn isotopes have been calculated by implementing of the new formula in Hartree-Fock-Bogoliubov (HFB) calculations. The results of the study shows that the new formula is useful for describing nuclear charge radii.
Artificial Neural Networks as a Tool for Galaxy Classification
Ofer Lahav
1996-12-10T23:59:59.000Z
We describe an Artificial Neural Network (ANN) approach to classification of galaxy images and spectra. ANNs can replicate the classification of galaxy images by a human expert to the same degree of agreement as that between two human experts, to within 2 T-type units. Similar methods are applied to classification of galaxy spectra. In particular, Principal Component Analysis of galaxy spectra can be used to compress the data, to suppress noise and to provide input to the ANNs. These and other classification methods will soon be applied to the Anglo-Australian 2-degree-Field (2dF) redshift survey of 250,000 galaxies.
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)
2014-07-14T23:59:59.000Z
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.
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)
2012-07-01T23:59:59.000Z
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)
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)
2005-04-09T23:59:59.000Z
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.
Mellit, A; Hadj-Arab, A; Guessoum, A
2004-01-01T23:59:59.000Z
An adaptive artificial neural network model for sizing stand-alone photovoltaic systems: Application for isolated sites in Algeria
, Sweden (To appear in Proceedings of the 7th ECNDT ¢¡¤£¦¥¨§©¥ In this paper we present a neural network
Neural Networks and Expert Systems to solve the problems of large amounts of Experimental Data at JET
Phase synchronization and chaotic dynamics in Hebbian learned artificial recurrent neural networks
Molter, Colin
Phase synchronization and chaotic dynamics in Hebbian learned artificial recurrent neural networks: increasing the storing capacity of recurrent neural networks as much as possible and observing and studying Colin Molter, Utku Salihoglu and Hugues Bersini Laboratory of artificial intelligence IRIDIA cp194
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
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
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
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
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
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
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
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
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
Sliding Mode Adaptive Neural-network Control for Nonholonomic Mobile Modular Manipulators
Li, Yangmin
Sliding Mode Adaptive Neural-network Control for Nonholonomic Mobile Modular Manipulators Yangmin adaptive neural-network controller for trajectory following of nonholonomic mobile modular manipulators model of the mobile modular manipulator. Sliding mode control and direct adaptive technique are combined
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
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))
1989-01-01T23:59:59.000Z
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.
Peterson, Carsten
Artificial Neural Networks and Human Expert for the Electrocardiographic Diagnosis of Healed Myocardial Infarction Hedén; Artificial Neural Network and Human Expert Bo Hedén, MD 1 , Mattias Ohlsson, PhD 2 , Ralf artificial neural networks and an experienced electrocardiographer diagnosing healed myocardial infarction
Dawson, Michael
How might an artificial neural network represent metric space? Patricia M.Boechler and Michael R.W.Dawson Department of Psychology,University of Alberta CONCLUS IONS Although this artificial neural network. + + + + + + + ABS TR ACT An artificial neural network was trained to rate the distances between pairs of cities
Bax, Ad
of an artificial neural network Yang Shen · Ad Bax Received: 4 June 2010 / Accepted: 30 June 2010 / Published a new chemical shift prediction program, SPARTA?, which is based on artificial neural networking. The neural network is trained on a large carefully pruned database, containing 580 proteins for which high
Sontag, Eduardo
Reprinted from: Artificial Neural Networks for Speech and Vision (Proc. Workshop held at Rutgers-125 UNIQUENESS OF WEIGHTS FOR NEURAL NETWORKS Francesca Albertini and Eduardo D. Sontag Department of Mathematics by neural networks have very little redundancy.) In this short expository survey, we sketch various known
Nielsen, Finn Årup
Artificial Neural Network Model for fMRI timeseries and a Framework for Comparison of Convolution of the hemodynamic response was proposed in [5]. Here we propose an artificial neural network (ANN) model. Method The specific neural network we will use is a twolayer feedforward type. Parameters are optimized
Zhuang, Qianlai
artificial neural networks Xudong Zhu,1 Qianlai Zhuang,1 Zhangcai Qin,1 Mikhail Glagolev,2 and Lulu Song1 develop a statistical model of CH4 emissions using an artificial neural network (ANN) approach and field to 2009 using artificial neural networks, Global Biogeochem. Cycles, 27, doi:10.1002/gbc.20052. 1
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
On the deduction of galaxy abundances with evolutionary neural networks
Michael Taylor; Angeles I. Diaz
2007-09-19T23:59:59.000Z
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*.
Smith, Alice E.
Estimation of All-Terminal Network Reliability Using an Artificial Neural Network Chat Srivaree be forwarded) Submitted to Computers & Operations Research June 1999 #12;1 Estimation of All-Terminal Network or communications network topologies, a common reliability measure is all-terminal reliability, the probability
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)
1991-12-01T23:59:59.000Z
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.
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)
2010-09-01T23:59:59.000Z
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.
Yao, Kun
2015-01-01T23:59:59.000Z
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.
An evaluation of neural networks for identification of system parameters in reactor noise signals
Miller, L.F.
1991-12-31T23:59:59.000Z
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.
1991-01-01T23:59:59.000Z
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.
Artificial Neural Network for search for metal poor galaxies
Shi, F; Kong, X; Chen, Y
2013-01-01T23:59:59.000Z
In order to find a fast and reliable method for selecting metal poor galaxies (MPGs), especially in large surveys and huge database, an Artificial Neural Network (ANN) method is applied to a sample of star-forming galaxies from the Sloan Digital Sky Survey (SDSS) data release 9 (DR9) provided by the Max Planck Institute and the Johns Hopkins University (MPA/JHU). A two-step approach is adopted:(i) The ANN network must be trained with a subset of objects that are known to be either MPGs or MRGs(Metal Rich galaxies), treating the strong emission line flux measurements as input feature vectors in an n-dimensional space, where n is the number of strong emission line flux ratios. (ii) After the network is trained on a sample of star-forming galaxies, remaining galaxies are classified in the automatic test analysis as either MPGs or MRGs. We consider several random divisions of the data into training and testing sets: for instance, for our sample, a total of 70 percent of the data are involved in training the algor...
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)
2008-06-12T23:59:59.000Z
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)
2009-09-09T23:59:59.000Z
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.
Galaxy Classification by Human Eyes and by Artificial Neural Networks
Ofer Lahav
1995-05-19T23:59:59.000Z
The rapid increase in data on galaxy images at low and high redshift calls for re-examination of the classification schemes and for new automatic objective methods. Here we present a classification method by Artificial Neural Networks. We also show results from a comparative study we carried out using a new sample of 830 APM digitised galaxy images. These galaxy images were classified by 6 experts independently. It is shown that the ANNs can replicate the classification by a human expert almost to the same degree of agreement as that between two human experts, to within 2 $T$-type units. Similar methods can be applied to automatic classification of galaxy spectra. We illustrate it by Principal Component Analysis of galaxy spectra, and discuss future large surveys.
ANNz: estimating photometric redshifts using artificial neural networks
Adrian A. Collister; Ofer Lahav
2004-02-27T23:59:59.000Z
We introduce ANNz, a freely available software package for photometric redshift estimation using Artificial Neural Networks. ANNz learns the relation between photometry and redshift from an appropriate training set of galaxies for which the redshift is already known. Where a large and representative training set is available ANNz is a highly competitive tool when compared with traditional template-fitting methods. The ANNz package is demonstrated on the Sloan Digital Sky Survey Data Release 1, and for this particular data set the r.m.s. redshift error in the range 0 < z < 0.7 is 0.023. Non-ideal conditions (spectroscopic sets which are small, or which are brighter than the photometric set for which redshifts are required) are simulated and the impact on the photometric redshift accuracy assessed.
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
1991-12-31T23:59:59.000Z
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)
1991-01-01T23:59:59.000Z
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.
NOVELTY DETECTION USING AUTO-ASSOCIATIVE NEURAL NETWORK
H. SOHN; K. WORDEN; C. FARRAR
2001-05-01T23:59:59.000Z
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
2011-08-30T23:59:59.000Z
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
2015-01-01T23:59:59.000Z
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
2014-12-17T23:59:59.000Z
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.
APPLICATION OF NEURAL NETWORK ALGORITHMS FOR BPM LINEARIZATION
Musson, John C. [JLAB; Seaton, Chad [JLAB; Spata, Mike F. [JLAB; Yan, Jianxun [JLAB
2012-11-01T23:59:59.000Z
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.
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)
2011-09-15T23:59:59.000Z
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.
Kansas, University of
Applications of Artificial Neural Networks and Fuzzy Models in High Throughput Screening to the existing HTS method, via Quantitative Structure-Activity Relationship (QSAR) using Artificial Neural in solving non-linear pattern classification problems, we propose several different models of neural networks
Application of artificial neural networks for damage indices classification with the use of Lamb of view. Artificial neural network has been used for the classification of fatigue cracks and artificial@agh.edu.pl, *corresponding author Keywords: NDT, Ultrasonic testing, Lamb waves, Artificial intelligence, Artificial Neural
Cambridge, University of
temperatures, as a function of chemical composition and heat treatments. Specification Language: FORTRAN / C The modelling procedure is a purely empirical one, and is based on a neural network program called generate44, which was developed by David MacKay and is part of the bigback5 program. The model is constituted
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
Singh, Harkirat
2004-09-30T23:59:59.000Z
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 systems. The AANN can...
Mining customer credit by using neural network model with logistic regression approach
Kao, Ling-Jing
2001-01-01T23:59:59.000Z
. 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...
A neural network mode inference engine for the advisory system for training and safety
Nguyen, Thinh Xuan
1996-01-01T23:59:59.000Z
logic membership functions. Although functional, the limitations of this method have prompted the development of an artificial neural network based SR (ANNSR). The goal of the ANNSR was to provide more accurate mode inferences, particularly during off...
Zeng, Xiaosi
2011-02-22T23:59:59.000Z
The artificial neural network (ANN) approach has been recognized as a capable technique to model the highly complex and nonlinear problem of travel time prediction. In addition to the nonlinearity, a traffic system is also temporally and spatially...
Analysis and forecast of the capesize bulk carriers shipping market using Artificial Neural Networks
Voudris, Athanasios V
2006-01-01T23:59:59.000Z
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 ...
An Analysis Method for Operations of Hot Water Heaters by Artificial Neural Networks
Yamaha, M.; Takahashi, M.
2004-01-01T23:59:59.000Z
Authors tried to apply an Artificial Neural Network (ANN) to estimation of state of building systems. The systems used in this study were gas combustion water heaters. Empirical equations to estimate gas consumption from measureble properies...
Long, Xiao
2012-10-19T23:59:59.000Z
performance of moment frames. Without such a model, the aforementioned problem can never be resolved. This dissertation develops an innovative way of predicting cyclic rupture in steel moment frames by employing artificial neural networks. First, finite...
Development of neural network calibration algorithms for multi-port pressure probes
Kinser Robert Eric
1996-01-01T23:59:59.000Z
The development of an enhanced neural network training algorithm and program for the calibration of velocity measurement instrumentation is presented. The backpropagation-based program, PROBENET, is intended to be a robust self learning code...
Prediction of MPEG-coded video source traffic using neural networks
Bhattacharya, Aninda
2002-01-01T23:59:59.000Z
with some success. This research demonstrates that video source traffic encoded using MPEG standards can be predicted single-step-ahead with reasonable amount of accuracy using neural networks as predictors. Use of two-step-ahead and four...
Mining customer credit by using neural network model with logistic regression approach
Kao, Ling-Jing
2001-01-01T23:59:59.000Z
. 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...
Data driven process monitoring based on neural networks and classification trees
Zhou, Yifeng
2005-11-01T23:59:59.000Z
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...
Comprehensive functional testing and dynamic compensation techniques for Cellular Neural Networks
Grimaila, Michael Russell
1995-01-01T23:59:59.000Z
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 ...
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)
2014-07-10T23:59:59.000Z
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.
PkANN: Non-Linear Matter Power Spectrum Interpolation through Artificial Neural Networks
Agarwal, Shankar
2012-12-31T23:59:59.000Z
We investigate the interpolation of power spectra of matter fluctuations using artificial neural networks (ANNs). We present a new approach to confront small-scale non-linearities in the matter power spectrum. This ...
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
1996-12-01T23:59:59.000Z
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.
Chen, X.; Liu, X.; Gales, M. J. F.; Woodland, P. C.
2015-04-22T23:59:59.000Z
, 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...
Incipient fault detection and identification in process systems using artificial neural networks
Muthusami, Jayakumar
1992-01-01T23:59:59.000Z
INCIPIENT FAULT DETECTION AND IDENTIFICATION IN PROCESS SYSTEMS USING ARTIFICIAL NEURAL NETWORKS A Thesis by JAYAKUMAR MUTHUSAMI 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 December 1992 Major Subject: Mechanical Engineering INCIPIENT FAULT DETECTION AND IDENTIFICATION IN PROCESS SYSTEMS USING ARTIFICIAL NEURAL NETWORKS A Thesis JAYAKHMAH MUTHTJSAMI Approved a. s to style and content, by...
Use of neural networks to identify transient operating conditions in nuclear power plants
Uhrig, R.E.; Guo, Z.
1989-01-01T23:59:59.000Z
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.
Nuclear mass systematics by complementing the Finite Range Droplet Model with neural networks
S. Athanassopoulos; E. Mavrommatis; K. A. Gernoth; J. W. Clark
2005-11-30T23:59:59.000Z
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.
Chang, Joongho
1994-01-01T23:59:59.000Z
A HIERARCHICAL STRUCTURE OF NEURAL NETWORK IMPLEMENTED FOR THE RECOGNITION OF AUTOMOBILE LICENSE PLATE NURSER A Thesis by JOONGHO CHANG Submitted to the Office of Graduate Studies of Texas A&M University in partial fulfdlment... of the requirements for the degree of MASTER OF SCIENCE August 1994 Major Subject: Electrical Engineering A HIERARCHICAL STRUCTURE OF NEURAL NETWORK IMPLEMENTED FOR THE RECOGNITION OF AUTOMOBILE LICENSE PLATE NUMBER A Thesis by JOONGHO CHANG Submitted...
Neural network technology for automatic fracture detection in sonic borehole image data
Schnorrenberg, Frank Theo
1992-01-01T23:59:59.000Z
NEURAL NETWORK TECHNOLOGY FOR AUTOMATIC FRACTURE DETECTION IN SONIC BOREHOLE IMAGE DATA A Thesis by FRANK THEO SCHNORRENBERG Submitted to the Office of Graduate Studies of Texas A&M University tn partial fulfillment of the requirements... for the degree of MASTER OF SC1ENCE December 1992 Major Subject: Computer Science NEURAL NETWORK TECHNOLOGY FOR AUTOMATIC FRACTURE DETECTION IN SONIC BOREHOLE IMAGE DATA A Thesis by FRANK THEO SCHNORRENBERG Subnutted to the Office of Graduate Studies...
Kischell, Eric Robert
1993-01-01T23:59:59.000Z
' (Member) A. D. Patton ( ead of epartment) August 1993 Major Subject: Electrical Engineering ABSTRACT Classification of Brain Compartments and Head Injury Lesions by Neural Networks Applied to Magnetic Resonance Images. (August 1993) Eric Robert... Kischell, B. S. , Northeastern University Chair of Advisory Committee: Dr. Nasser Kehtarnavaz An automatic neural network-based approach was ap- plied to segment brain compartments and closed-head-injury lesions in magnetic resonance (MR) images Two...
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
1996-11-01T23:59:59.000Z
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.
Sontag, Eduardo
Reprinted from: Artificial Neural Networks for Speech and Vision (Proc. Workshop held at Rutgers125 UNIQUENESS OF WEIGHTS FOR NEURAL NETWORKS Francesca Albertini \\Lambda and Eduardo D. Sontag y Department into a network, in order to increase representational bias, by imposing artificial conditions such as asking
Automatic Detection of Expanding HI Shells Using Artificial Neural Networks
Anik Daigle; Gilles Joncas; Marc Parizeau; Marc-Antoine Miville-Deschenes
2003-04-17T23:59:59.000Z
The identification of expanding HI shells is difficult because of their variable morphological characteristics. The detection of HI bubbles on a global scale therefore never has been attempted. In this paper, an automatic detector for expanding HI shells is presented. The detection is based on the more stable dynamical characteristics of expanding shells and is performed in two stages. The first one is the recognition of the dynamical signature of an expanding bubble in the velocity spectra, based on the classification of an artificial neural network. The pixels associated with these recognized spectra are identified on each velocity channel. The second stage consists in looking for concentrations of those pixels that were firstly pointed out, and to decide if they are potential detections by morphological and 21-cm emission variation considerations. Two test bubbles are correctly detected and a potentially new case of shell that is visually very convincing is discovered. About 0.6% of the surveyed pixels are identified as part of a bubble. These may be false detections, but still constitute regions of space with high probability of finding an expanding shell. The subsequent search field is thus significantly reduced. We intend to conduct in the near future a large scale HI shells detection over the Perseus Arm using our detector.
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.
1993-02-23T23:59:59.000Z
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., E-mail: meneghini@fusion.gat.com [Oak Ridge Associated Universities, 120 Badger Ave, Oak Ridge, Tennessee 37830 (United States); Luna, C. J. [Arizona State University, 411 N. Central Ave, Phoenix, Arizona 85004 (United States); Smith, S. P.; Lao, L. L. [General Atomics, San Diego, California 92186-5608 (United States)
2014-06-15T23:59:59.000Z
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.
Multi-parameter estimating photometric redshifts with artificial neural networks
Lili Li; Yanxia Zhang; Yongheng Zhao; Dawei Yang
2007-04-17T23:59:59.000Z
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.
2007-08-01T23:59:59.000Z
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.
, Sweden (To appear in Proceedings of the 7 th ECNDT 6i+#...hp# In this paper we present a neural network
Zuo Guangqing; Ma Jitang; Bo, B. [Luleaa Univ. of Technology (Sweden). Div. of Process Metallurgy
1996-12-31T23:59:59.000Z
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)
2010-08-15T23:59:59.000Z
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.
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
Neurocomputing 70 (2006) 603606 Stability analysis of an unsupervised neural network
Pilyugin, Sergei S.
2006-01-01T23:59:59.000Z
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))
1993-02-01T23:59:59.000Z
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
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)
2002-07-01T23:59:59.000Z
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)
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)
2010-02-22T23:59:59.000Z
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.
Sridhar, Srinivas
component and artificial neural network analysis Susie Boydston-White a,b , Melissa Romeo a,b , Tatyana.V. All rights reserved. Keywords: Cell cycle; Infrared micro-spectroscopy; Artificial neural network
Uhrig, R.E. (Tennessee Univ., Knoxville, TN (United States) Oak Ridge National Lab., TN (United States))
1990-01-01T23:59:59.000Z
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))
1990-01-01T23:59:59.000Z
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.
Cambridge, University of
30 8. Neural Networks Over the years, linear regression models have attempted to characterise the 0 interact. A more powerful alternative is the use of neural networks [40,42], a non-linear modelling prediction uncertainties. #12;31 In linear regression, the sum of each input xi multiplied with a weight wi
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
Gupta, Vinay Kumar
28 ARTIFICIAL NEURAL NETWORK-BASED ESTIMATION OF PEAK GROUND ACCELERATION C.R. Arjun and Ashok Kumar Department the application of artificial neural networks (ANNs) for the estimation of peak ground acceleration (PGA inputs. KEYWORDS: Artificial Neural Networks, Peak Ground Acceleration, Hypocentral Distance, Shear Wave
Paris-Sud XI, Université de
99-72 Application of an Artificial Neural Network to the Prediction of Firedamp Emissions in Coal-linear physical laws and a high number of hardly accessible parameters. So artificial neural networks have been developed to model firedamp emissions : artificial neural networks äs universal approx'imators are able
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
Systematic uncertainties of artificial neural-network pulse-shape discrimination for $0\
Abt, I; Cossavella, F; Majorovits, B; Palioselitis, D; Volynets, O
2014-01-01T23:59:59.000Z
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 the systematic uncertainties of the method. 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 samples from calibration measurements is estimated to be 5\\%. This uncertainty is due to differences between signal and calibration samples.
Application of neural networking in live cattle futures market: an approach to price-forecasting
Chou, Chien-Ju
1993-01-01T23:59:59.000Z
APPLICATION OF NEURAL NETWORKING IN LIVE CATTLE FUTURES MARKET: AN APPROACH TO PRICE-FORECASTING A Thesis by CHIEN-JU CHOU Submitted to the Office of Graduate Studies of Texas ARM University in partial fulfilhnent of the requirements... for the degree of MASTER OF SCIENCE August 1993 Major Subject: Animal Science APPLICATION OF NEURAL NETWORKING IN LIVE CATTLE FUTURES MARKET: AN APPROACH TO PRICE-FORECASTING A Thesis by CHIBN-JU CHOU Approved as to style and content by: John P. Walter...
Development of Ensemble Neural Network Convection Parameterizations for Climate Models
Fox-Rabinovitz, M. S.; Krasnopolsky, V. M.
2012-05-02T23:59:59.000Z
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)
2014-07-10T23:59:59.000Z
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.
Coutinho, Alvaro L. G. A.
the viscoelastic parameters; 3D FEM analysis using High Performance Computing (parallel and vector features) to run Performance Computing. E. M. R. Fairbairn, E. Goulart, A. L. G. A. Coutinho, N. F. F. Ebecken COPPEDurability Assessment of an Arch Dam using Inverse Analysis with Neural Networks and High
Cottrell, Garrison W.
facial expressions is compared with human subjects over a set of experiments using interpolated imagery. The experiments for both the human subjects and neural networks make use of in terpolations of facial expressions in materials between those used in the human subjects experiments [Young et al., 1997] and our materials
Energy Management System for an Hybrid Electric Vehicle, Using Ultracapacitors and Neural Networks
Catholic University of Chile (Universidad Católica de Chile)
Energy Management System for an Hybrid Electric Vehicle, Using Ultracapacitors and Neural Networks and specific energy contained in most electric batteries compared to that of gasoline, is resolved in hybrid to accept energy from regenerative braking. For this reason, hybrid systems use an auxiliary energy system
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
Weiss, Gary
Utility based Data Mining for Time Series Analysis - Cost-sensitive Learning for Neural Network@bis-lab.com ABSTRACT In corporate data mining applications, cost-sensitive learning is firmly established Mining General Terms Algorithms, Management, Economics Keywords Data Mining, cost-sensitive learning
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
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
Designing Polymer Blends Using Neural Networks, Genetic Algorithms, and Markov Chains
Potter, Don
1 Designing Polymer Blends Using Neural Networks, Genetic Algorithms, and Markov Chains N. K. Roy1 we present a new technique to simulate polymer blends that overcomes the shortcomings in polymer system modeling. This method has an inherent advantage in that the vast existing information on polymer
Network: Computation in Neural Systems March 2007; 18(1): 13
Shadmehr, Reza
Network: Computation in Neural Systems March 2007; 18(1): 13 BOOK REVIEW The Computational and systems-level computational neuroscience. Low-level computational neuroscience is primarily concerned: 10.1080/09548980701275714 #12;computational neuroscience looks more into how the nervous system
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
Forecasting of preprocessed daily solar radiation time series using neural networks
Boyer, Edmond
Forecasting of preprocessed daily solar radiation time series using neural networks Christophe prediction of global solar radiation on a horizontal surface. First results are promising with nRMSE ~ 21 t or at day d and year y d H0 Extraterrestrial solar radiation coefficient for day d [MJ/m²] xt, xd,y Time
Annan, Carl Ashie
2014-12-18T23:59:59.000Z
(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...
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
Neural Network-Based Classification of Single-Phase Distribution Transformer Fault Data
Zhang, Xujia
2006-08-16T23:59:59.000Z
an efficient and accurate classification method to separate monitoring signal data into clusters base on their properties. However, raw data collected from the field and simulations will create too many dimensions and inputs to the neural network and make it a...
Cambridge, University of
on experience and test results of creep data obtained after long-term creep tests for several years or moreCREEP STRENGTH OF HIGH CR FERRITIC STEELS DESIGNED USING NEURAL NETWORKS AND PHASE STABILITY of Materials Science and Metallurgy Pembroke Street, Cambridge CB2 3QZ, U.K. Abstract The highest creep rupture
HotStrength of Ferritic CreepResistant Steels Comparison of Neural Network and Genetic Programming
Cambridge, University of
recently analysed how the strength of creepresistant ferritic steels depends on the testrupture provides a ready method for estimating expensive creep data using ordinary tensile tests [1]. AnotherHotStrength of Ferritic CreepResistant Steels Comparison of Neural Network and Genetic
Neural NetworkBased Modeling and Optimization for Effective Vehicle Emission Testing and
Huang, Yinlun
Neural NetworkBased Modeling and Optimization for Effective Vehicle Emission Testing and Engine and Materials Science, Wayne State University, Detroit, Michigan, USA In automotive manufacturing, vehicle emission testing and engine calibration are the key to achieving emission standards with satisfactory fuel
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
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
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
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
Reservoir characterization using seismic attributes, well data, and artificial neural networks
Toinet, Sylvain
2001-01-01T23:59:59.000Z
. 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...
Prediction of Full-Scale Propulsion Power using Artificial Neural Networks
@imm.dtu.dk Abstract Full scale measurements of the propulsion power, ship speed, wind speed and direction, sea and air on the measurements of a set of significant input variables, which are: · Ship speed through the water · Wind speed537 Prediction of Full-Scale Propulsion Power using Artificial Neural Networks Benjamin Pjedsted
Annan, Carl Ashie
2014-12-18T23:59:59.000Z
. This work proposes a neural network approach to quantify the degree of stiction in a valve once the phenomenon has been detected. Several degrees of stiction are simulated in a closed loop control system by specifying the magnitude of static (fs) and dynamic...
Zhang, Xuesong; Liang, Faming; Yu, Beibei; Zong, Ziliang
2011-11-09T23:59:59.000Z
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.
Modifed Minimum Classification Error Learning and Its Application to Neural Networks
Shimodaira, Hiroshi; Rokui, Jun; Nakai, Mitsuru
-fitting" to the training samples. In the present study, a regularization technique has been employed to the MCE learning to overcome this problem. Feed-forward neural networks are employed as a recognition platform to evaluate the recognition performance of the proposed...
Neural networks for modelling the final target cost of water projects
Ahiaga-Dagbui, Dominic D; Smith, Simon D
capabilities of Artificial Neural Networks (ANN) are explored in this pilot study to build final cost estimation models that incorporate the cost effect of some of the factors mentioned above. Data was collected on ninety-eight water-related construction...
Creation and Testing of an Artificial Neural Network Based Carbonate Detector for Mars Rovers
Royer, Dana
carbonate detector capable of running on current and future rover hardware. The detector can identify result in a critical loss of valuable science data. Onboard techniques for identifying 1 0 neural network (ANN) based carbonate detector capable of running on current and future rover hardware
JPEG Quality Transcoding using Neural Networks Trained with a Perceptual Error Measure
Lazzaro, John
JPEG Quality Transcoding using Neural Networks Trained with a Perceptual Error Measure John Lazzaro@cs.berkeley.edu Abstract A JPEG Quality Transcoder (JQT) converts a JPEG image file that was encoded with low image quality users direct control over the compression process, supporting trade- offs between image quality
Ang Jr.,, Marcelo H.
1 Neural Network Controller for Constrained Robot Manipulators Shenghai Hu, Marcelo H. Ang Jr robot manipulators to a wider class of tasks, it is necessary to control not only the position of a manipulator but also the force exerted by its end-effector on an object or environment. Force control
Sliding Mode Adaptive Neural-Network Control for Nonholonomic Mobile Modular Manipulators
Li, Yangmin
Sliding Mode Adaptive Neural-Network Control for Nonholonomic Mobile Modular Manipulators YUGANG the dynamic model of the mobile mod- ular manipulator. Sliding mode control and direct adaptive technique modular manipulator, sliding mode control. 1. Introduction In the past decades, modular manipulators have
Massachusetts at Amherst, University of
Decentralized kinematic control of a class of collaborative redundant manipulators via recurrent neural network Quadratic programming Cooperative task execution Redundant manipulator Decentralized kinematic control a b s t r a c t This paper studies the decentralized kinematic control of multiple
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
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
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