A fuzzy-controlled delta-bar-delta learning rule

Author

Lippe, W.-M. ; Feuring, Th ; Tenhagen, A.

Author_Institution

Inst. fur Num. Math./Inf., Westfalischen Wilhelms-Univ., Munster, Germany

Volume

3

fYear

1994

fDate

27 Jun-2 Jul 1994

Firstpage

1686

Abstract

In classic backpropagation nets, as introduced by Rumelhart et al. (1986), the weights are modified according to the method of steepest descent. The goal of this weight modification is to minimise the error in net-outputs for a given training set. Basing upon Jacobs´ work (1988), we point out drawbacks of steepest descent and suggest improvements on it. These yield a backpropagation net, which adjusts its weights according to a parallel coordinate descent method, whose parameters are being fuzzy-controlled

Keywords

backpropagation; fuzzy control; fuzzy neural nets; multilayer perceptrons; backpropagation nets; error minimisation; fuzzy parameter control; fuzzy-controlled delta-bar-delta learning rule; parallel coordinate descent method; steepest descent; weight modification; Approximation algorithms; Backpropagation algorithms; Convergence; Fuzzy control; Jacobian matrices; Neural networks; Newton method; Rough surfaces; Surface roughness; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 1994. IEEE World Congress on Computational Intelligence., 1994 IEEE International Conference on

Conference_Location

Orlando, FL

Print_ISBN

0-7803-1901-X

Type

conf

DOI

10.1109/ICNN.1994.374410

Filename

374410