DMCA
Optimization Strategies And Backpropagation Neural Networks
by
V. P. Plagianakos
,
G. D. Magoulas
,
M. N. Vrahatis
BibTeX
@MISC{Plagianakos_optimizationstrategies,
author = {V. P. Plagianakos and G. D. Magoulas and M. N. Vrahatis},
title = {Optimization Strategies And Backpropagation Neural Networks},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
this paper we have presented new learning strategies for generating monotone and nonmonotone learning rules for neural network training. The simulation results were satisfactory and suggest that the use of the proposed strategies, can signi#cantly accelerate the convergence of the training algorithms. Additionally, the use of di#cult to tune problem-dependent heuristic parameters is unnecessary; the modi#ed methods proved to be robust against phenomena such as oscillations due to poorly chosen heuristics parameters. References
