Open Access. Powered by Scholars. Published by Universities.®
Physical Sciences and Mathematics Commons™
Open Access. Powered by Scholars. Published by Universities.®
- Keyword
-
- Computation Times (1)
- Convergence (1)
- Godunov method (1)
- Intel IPSC/2 (1)
- Learning Rates (1)
-
- Learning Systems (1)
- Least Squares Approximations (1)
- Marquardt-Levenberg Least-Square Optimization Method (1)
- Neural Nets (1)
- Optimisation (1)
- Parallel Architectures (1)
- Parallel Processing (1)
- Partial differential equations (1)
- Recursive Least Squares Neural Network Training (1)
- Shock-capturing schemes (1)
- Supervised Learning (1)
- Publication
- Publication Type
Articles 1 - 3 of 3
Full-Text Articles in Physical Sciences and Mathematics
Parallel Implementation Of A Recursive Least Squares Neural Network Training Method On The Intel Ipsc/2, James Edward Steck, Bruce M. Mcmillin, K. Krishnamurthy, M. Reza Ashouri, Gary G. Leininger
Parallel Implementation Of A Recursive Least Squares Neural Network Training Method On The Intel Ipsc/2, James Edward Steck, Bruce M. Mcmillin, K. Krishnamurthy, M. Reza Ashouri, Gary G. Leininger
Computer Science Faculty Research & Creative Works
An algorithm based on the Marquardt-Levenberg least-square optimization method has been shown by S. Kollias and D. Anastassiou (IEEE Trans. on Circuits Syst. vol.36, no.8, p.1092-101, Aug. 1989) to be a much more efficient training method than gradient descent, when applied to some small feedforward neural networks. Yet, for many applications, the increase in computational complexity of the method outweighs any gain in learning rate obtained over current training methods. However, the least-squares method can be more efficiently implemented on parallel architectures than standard methods. This is demonstrated by comparing computation times and learning rates for the least-squares method implemented …
An Extension Of Essentially Non-Oscillatory Shock-Capturing Schemes To Multi-Dimensional Systems Of Conservation Laws, Jay Casper
Mathematics & Statistics Theses & Dissertations
In recent years, a class of numerical schemes for solving hyperbolic partial differential equations has been developed which generalizes the first-order method of Godunov to arbitrary order of accuracy. High-order accuracy is obtained, wherever the solution is smooth, by an essentially non-oscillatory (ENO) piecewise polynomial reconstruction procedure, which yields high-order pointwise information from the cell averages of the solution at a given point in time. When applied to piecewise smooth initial data, this reconstruction enables a flux computation that provides a time update of the solution which is of high-order accuracy, wherever the function is smooth, and avoids a Gibbs …
A Study Of The Hydration Properties Of Selected Laser Dye Aerosols Including Continuous-Flow Parallel Plate And Alternating-Gradient Thermal Diffusion Cloud Chamber Measurements In The High Supersaturation Regime, Donald E. Hagen, Max B. Trueblood, Darryl J. Alofs
A Study Of The Hydration Properties Of Selected Laser Dye Aerosols Including Continuous-Flow Parallel Plate And Alternating-Gradient Thermal Diffusion Cloud Chamber Measurements In The High Supersaturation Regime, Donald E. Hagen, Max B. Trueblood, Darryl J. Alofs
Mechanical and Aerospace Engineering Faculty Research & Creative Works
The Hydration Behavior of Aerosols made Up of Various Fluorescent Dyes when Exposed to Water Saturation or Supersaturated Conditions Has Been Studied. Critical Supersaturation Spectra Are Reported. the Dyes Are Found to Behave as High Molecular Weight Ionic Compounds that Obey Kohler Theory. their Relevant Kohler Parameters Are Measured. This Study Makes Use of and Compares Results from the Isothermal Hazecontinuous-Flowand Alternating-Gradient Thermal Diffusion Cloud Chambers. the Ability of the Continuous-Flow Thermal Diffusion Chamber to Operate Correctly at High Supersaturations is Shown. © 1990 Elsevier Science Publishing Co., Inc.