Presentation 2000/11/17
A Cooperative Learning Algorithm for Multi-Layered Neural Networks by using a PC Cluster
Eiji Watanabe, Katsumi Mori,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this report, a cooperative learning algorithm of multi-layered neural networks(NNs)is newly proposed based on a PC cluster and the experimental results for pattern classification problems are shown.The proposed cooperative learning algorithm has the following procedures : (i)parallel learning of each NN having different initial weights by using the error back propagation learning algorithm, (ii)exchange of information(weights)so as to minimize the unbalance among NNs and improve the learning performance of them.Finally, we apply the proposed cooperative learning algorithm for a classfication problem of Japanese-Kanji fonts and show its behaviour and effectiveness concretely.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Multi-layered neural network / PC cluster / Module structure / Cooperative learning algorithm / Information exchange
Paper # NC2000-71
Date of Issue

Conference Information
Committee NC
Conference Date 2000/11/17(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 Cooperative Learning Algorithm for Multi-Layered Neural Networks by using a PC Cluster
Sub Title (in English)
Keyword(1) Multi-layered neural network
Keyword(2) PC cluster
Keyword(3) Module structure
Keyword(4) Cooperative learning algorithm
Keyword(5) Information exchange
1st Author's Name Eiji Watanabe
1st Author's Affiliation Department of Information Systems and Management Science Faculty of Science, Konan University()
2nd Author's Name Katsumi Mori
2nd Author's Affiliation Department of Information Processing Engineering Faculty of Engineering, Fukuyama University
Date 2000/11/17
Paper # NC2000-71
Volume (vol) vol.100
Number (no) 466
Page pp.pp.-
#Pages 8
Date of Issue