Presentation 1997/3/17
Dynamics of Autoassociative Neural Networks
Satoshi Morinaga, Kazuhiro Ishida, Shuji Yoshizawa,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this report, the dynamics of autoassociative analogue neural networks is analyzed theoretically. First, we show that the distance between the network state vector and the subspace Π_1 spanned by given patterns decreases in exponential order of time. And we show that the network dynamics is a gradient flow in Π_1. Then we prove that the dynamics outside of Π_1 is also a gradient flow. It means that the dynamics is not chaotic. Finally, we look into the dynamics in Π_1 and the dynamics toward Π_1 by computer simulations.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) associative memory / analogue neural networks / dynamics / chaos / gradient flow
Paper # NC96-125
Date of Issue

Conference Information
Committee NC
Conference Date 1997/3/17(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) Dynamics of Autoassociative Neural Networks
Sub Title (in English)
Keyword(1) associative memory
Keyword(2) analogue neural networks
Keyword(3) dynamics
Keyword(4) chaos
Keyword(5) gradient flow
1st Author's Name Satoshi Morinaga
1st Author's Affiliation C&C Research Laboratories NEC Corporation()
2nd Author's Name Kazuhiro Ishida
2nd Author's Affiliation University of Tokyo
3rd Author's Name Shuji Yoshizawa
3rd Author's Affiliation University of Tokyo
Date 1997/3/17
Paper # NC96-125
Volume (vol) vol.96
Number (no) 583
Page pp.pp.-
#Pages 7
Date of Issue