Presentation 2003/11/15
Analysis of Input Power of Control of Assosiative Chaotic Neural Networks for Aperiodic Using Reinforcement Learning
Norihisa SATO, Masaharu ADACHI, Makoto KOTANI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) An anti-control of associative chaotic neural networks using reinforcement learning has been proposed and an example of successful control has been shown. However it has not been reported about actual control inputs in detail. Therefore in this article, we analyze time series of control inputs that are determined by the state-action value function obtained by the Q-learning. As a result, it is found that transition frequency among control input values tends to be uniformly distributed when the anti-control succeeds. On the other hand, the distribution tends to concentrate to certain combinations of the input values when it fails. However, we can not find any simple relationship between the learning time and success rate of the anti-control.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) anti-control of chaos / reinforcement lerning / chaotic neural networks
Paper # NLP2003-126
Date of Issue

Conference Information
Committee NLP
Conference Date 2003/11/15(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 Nonlinear Problems (NLP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Analysis of Input Power of Control of Assosiative Chaotic Neural Networks for Aperiodic Using Reinforcement Learning
Sub Title (in English)
Keyword(1) anti-control of chaos
Keyword(2) reinforcement lerning
Keyword(3) chaotic neural networks
1st Author's Name Norihisa SATO
1st Author's Affiliation Department of Electronic Engineering,Graduate School of Engineering. Tokyo Denki University()
2nd Author's Name Masaharu ADACHI
2nd Author's Affiliation Department of Electronic Engineering,Graduate School of Engineering. Tokyo Denki University
3rd Author's Name Makoto KOTANI
3rd Author's Affiliation Department of Electronic Engineering,Graduate School of Engineering. Tokyo Denki University
Date 2003/11/15
Paper # NLP2003-126
Volume (vol) vol.103
Number (no) 464
Page pp.pp.-
#Pages 6
Date of Issue