Presentation 1994/5/19
Educational Learning for Multi-layered Neural Network Appending Synaptic Weights to Performance Index
Tatsuya Morita, Kazuyoshi Nakajima, Masahiro Tamura,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The authors previously proposed a neural network model having a multi-layered structure similar to the butterfly processing of fast Fourier transformation.Also,a new learning method called educational learning was proposed instead of the back propagation method. In this paper,intermediary target production process in the educational learning is analyzed using optimization theory. Learning performance can be improved by appending terms of synaptic weights and thresholds to the performance index in the intermediary target production process. The theoretical aspect is examined by a simulation of pattern transformation problem.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Multi-layered neural network / Educational learning / Intermediate target production / Application of optimization theory / Pattern transformation problem
Paper # NC94-5
Date of Issue

Conference Information
Committee NC
Conference Date 1994/5/19(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) Educational Learning for Multi-layered Neural Network Appending Synaptic Weights to Performance Index
Sub Title (in English)
Keyword(1) Multi-layered neural network
Keyword(2) Educational learning
Keyword(3) Intermediate target production
Keyword(4) Application of optimization theory
Keyword(5) Pattern transformation problem
1st Author's Name Tatsuya Morita
1st Author's Affiliation Department of Electrical Engineering,Faculty of Engineering,Osaka University.()
2nd Author's Name Kazuyoshi Nakajima
2nd Author's Affiliation Department of Electrical Engineering,Faculty of Engineering,Osaka University.
3rd Author's Name Masahiro Tamura
3rd Author's Affiliation Department of Electrical Engineering,Faculty of Engineering,Osaka University.
Date 1994/5/19
Paper # NC94-5
Volume (vol) vol.94
Number (no) 40
Page pp.pp.-
#Pages 7
Date of Issue