Title :
Application of GRG2 for training neural networks
Author :
Subramanian, Venkat ; Hung, Ming S.
Author_Institution :
Wisconsin Univ., Kenosha, WI, USA
Abstract :
The authors present a new training system based on GRG2, a widely distributed nonlinear optimization software. Comparisons with backpropagation based upon three benchmark problems suggest not only that the GRG2-based system is much faster, robust, and offers solutions with better quality, but that it also offers better scalability to larger problems
Keywords :
computer aided instruction; distributed processing; learning systems; mathematics computing; neural nets; nonlinear programming; GRG2; Generalised Reduced Gradient; backpropagation; distributed nonlinear optimization software; encoding; neural networks; nonlinear programming; scalability; training system; Artificial neural networks; Convergence; Design optimization; Feedforward systems; Least squares approximation; Neural networks; Robustness; Scalability; Software algorithms; Spirals;
Conference_Titel :
Neural Networks, 1991., IJCNN-91-Seattle International Joint Conference on
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-0164-1
DOI :
10.1109/IJCNN.1991.155339
