On the effect of the nonlinearity of the sigmoid function in artificial neural network classifiers

Author

Uchimura, S. ; Hamamoto, Yoshihiko ; Tomita, Shingo

Author_Institution

Oshima Nat. Coll. of Maritime Technol., Japan

Volume

1

fYear

1995

fDate

Nov/Dec 1995

Firstpage

281

Abstract

Despite a considerable amount of recent directed towards pattern recognition applications of the artificial neural networks (ANNs), little quantitative information concerning the nonlinearity of the sigmoid function is available to a designer. In this paper, the authors study the effect of the nonlinearity of the sigmoid function on the generalization capability of the ANN classifiers trained with the backpropagation algorithm. Experimental results show that like the case of the network size, there exists the optimal degree of the nonlinearity for a particular problem

Keywords

backpropagation; neural nets; pattern classification; artificial neural network classifiers; backpropagation algorithm; generalization capability; network size; nonlinearity; pattern recognition; sigmoid function; Artificial neural networks; Backpropagation algorithms; Classification algorithms; Convergence; Design engineering; Educational institutions; Electronic mail; Intelligent networks; Linearity; Neurons; Pattern recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 1995. Proceedings., IEEE International Conference on

Conference_Location

Perth, WA

Print_ISBN

0-7803-2768-3

Type

conf

DOI

10.1109/ICNN.1995.488109

Filename

488109