A New Stochastic PSO Technique for Neural Network Training Yangmin Li and Xin Chen Department Optimization(PSO) has been widely applied for training neural network. To improve the performance of PSO paradigm of particle swarm optimization named stochastic PSO (S-PSO). The feature of the S-PSO is its high
Procedures (CHP) `model' produced by the SPSO and, where appropriate, reflects the guidance laid out in that CHP. All HEIs are required by the Public Services Reform (Scotland) Act 2010 to comply with the CHP
.mo Abstract-- This paper proposes a new approach using im- proved particle swarm optimization (PSO optimization methods that produce nonsmooth paths, the PSO method developed in this paper can generate smooth using mobile robots. To reduce computational cost of optimization, the stochastic PSO (S-PSO) with high
Neural Network Training Using Stochastic PSO Xin Chen and Yangmin Li Department is huge, when PSO algorithms are applied for NN training, the dimension of search space is so large that PSOs always converge prematurely. In this paper an improved stochastic PSO (SPSO) is presented
Gajalakshmi, Palanivel; Priya, Mani Krishna; Pradeep, Thangaraj; Behera, Jyotirmaya; Muthumani, Kandasamy; Madhuwanti, Srinivasan; Saran, Uttara; Chatterjee, Suvro
Widely used chemotherapeutic breast cancer drugs such as Tamoxifen citrate (TC), Capecitabine (CP) and Epirubicin (EP) are known to cause various cardiovascular side-effects among long term cancer survivors. Vascular modulation warrants nitric oxide (NO) signal transduction, which targets the vascular endothelium. We hypothesize that TC, CP and EP interference with the nitric oxide downstream signaling specifically, could lead to cardiovascular dysfunctions. The results demonstrate that while all three drugs attenuate NO and cyclic guanosine mono-phosphate (cGMP) production in endothelial cells, they caused elevated levels of NO in the plasma and RBC. However, PBMC and platelets did not show any significant changes under treatment. This implies that the drug effects are specific to the endothelium. Altered eNOS and phosphorylated eNOS (Ser-1177) localization patterns in endothelial cells were observed following drug treatments. Similarly, the expression of phosphorylated eNOS (Ser-1177) protein was decreased under the treatment of drugs. Altered actin polymerization was also observed following drug treatment, while addition of SpNO and 8Br-cGMP reversed this effect. Incubation with the drugs decreased endothelial cell migration whereas addition of YC-1, SC and 8Br-cGMP recovered the effect. Additionally molecular docking studies showed that all three drugs exhibited a strong binding affinity with the catalytic domain of human sGC. In conclusion, results indicate that TC, CP and EP cause endothelial dysfunctions via the NO–sGC–cGMP pathway and these effects could be recovered using pharmaceutical agonists of NO signaling pathway. Further, the study proposes a combination therapy of chemotherapeutic drugs and cGMP analogs, which would confer protection against chemotherapy mediated vascular dysfunctions in cancer patients. - Highlights: • NO production is reduced in endothelial cells under breast cancer drug treatment. • Cellular cGMP level is decreased under the treatments of breast cancer drugs. • Breast cancer drugs induce vasoconstriction by interfering with NO pathway. • NO donors, cGMP analogs rescue breast cancer drug induced endothelial dysfunctions.
Larry G. Stolarczyk
With the aid of a DOE grant (No. DE-FC26-01NT41050), Stolar Research Corporation (Stolar) developed the Horizon Sensor (HS) to distinguish between the different layers of a coal seam. Mounted on mining machine cutter drums, HS units can detect or sense the horizon between the coal seam and the roof and floor rock, providing the opportunity to accurately mine the section of the seam most desired. HS also enables accurate cutting of minimum height if that is the operator's objective. Often when cutting is done out-of-seam, the head-positioning function facilitates a fixed mining height to minimize dilution. With this technology, miners can still be at a remote location, yet cut only the clean coal, resulting in a much more efficient overall process. The objectives of this project were to demonstrate the feasibility of horizon sensing on mining machines and demonstrate that Horizon Sensing can allow coal to be cut cleaner and more efficiently. Stolar's primary goal was to develop the Horizon Sensor (HS) into an enabling technology for full or partial automation or ''agile mining''. This technical innovation (R&D 100 Award Winner) is quickly demonstrating improvements in productivity and miner safety at several prominent coal mines in the United States. In addition, the HS system can enable the cutting of cleaner coal. Stolar has driven the HS program on the philosophy that cutting cleaner coal means burning cleaner coal. The sensor, located inches from the cutting bits, is based upon the physics principles of a Resonant Microstrip Patch Antenna (RMPA). When it is in proximity of the rock-coal interface, the RMPA impedance varies depending on the thickness of uncut coal. The impedance is measured by the computer-controlled electronics and then sent by radio waves to the mining machine. The worker at the machine can read the data via a Graphical User Interface, displaying a color-coded image of the coal being cut, and direct the machine appropriately. The Horizon Sensor program began development in 1998 and experienced three major design phases. The final version, termed HS-3, was commissioned in 2000 with the assistance of the DOE-Mining Industry of the Future program, commercialized in 2002, and has been used 14 times in 12 different mines within the United States. The Horizon Sensor has applications in both underground and surface mining operations. This technology is primarily used in the coal industry, but is also used to mine trona and potash. All horizon sensor components have Mine Safety and Health Administration (MSHA) (United States) and IEC (International) certification. Horizon Sensing saves energy by maximizing cutting efficiency, cutting only desired material. This desired material is cleaner fuel, therefore reducing pollutants to the atmosphere when burned and burning more efficiently. Extracting only desired material increases productivity by reducing or eliminating the cleaning step after extraction. Additionally, this technology allows for deeper mining, resulting in more material gained from one location. The remote sensing tool allows workers to operate the machinery away from the hazards of cutting coal, including noise, breathing dust and gases, and coal and rock splintering and outbursts. The HS program has primarily revolved around the development of the technology. However, the end goal of the program has always been the commercialization of the technology and only within the last 2 years of the program has this goal been realized. Real-time horizon sensing on mining machines is becoming an industry tool. Detailed monitoring of system function, user experience, and mining benefits is ongoing.