Presentation 1997/3/18
A Modular Neural Network using Fuzzy Competitive Learning
Hirotsugu Kikuakawa, Takashi Nagano,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this study, We Propose a new modular neural network using FCL(Fuzzy Competitive Learning). The FCL makes a gating network give an appropriate space division independently of the learning of expert's modules. Morever we propose the pruning algorithm to determine appropriate architecture of the model. This algorithm makes the model obtain minimum number of expert networks and the maximum performance, because its procedure estimates the contribution ratio of each expert network and then if a contribution ratio is less than an estimation criterion, it is pruned from the whole network. The performance of the network was examined by computer simulations.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) FCL / a modular neural network / pruning algorithm
Paper # NC96-159
Date of Issue

Conference Information
Committee NC
Conference Date 1997/3/18(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Modular Neural Network using Fuzzy Competitive Learning
Sub Title (in English)
Keyword(1) FCL
Keyword(2) a modular neural network
Keyword(3) pruning algorithm
1st Author's Name Hirotsugu Kikuakawa
1st Author's Affiliation Department of Industrial and System Engineering,College of Engineering,Hosei University()
2nd Author's Name Takashi Nagano
2nd Author's Affiliation Department of Industrial and System Engineering,College of Engineering,Hosei University
Date 1997/3/18
Paper # NC96-159
Volume (vol) vol.96
Number (no) 584
Page pp.pp.-
#Pages 8
Date of Issue