Summary

International Symposium on Nonlinear Theory and its Applications

2005

Session Number:4-3-1

Session:

Number:4-3-1-3

Influence of Randomly Connecting External Inputs on Associative Chaotic Neurodynamics

Masaharu ADACHI,  Kazuyuki AIHARA,  

pp.50-53

Publication Date:2005/10/18

Online ISSN:2188-5079

DOI:10.34385/proc.40.4-3-1-3

PDF download (93.3KB)

Summary:
In the present paper, we investigate influence of randomly connecting external inputs on retrievals in associative chaotic neural networks. The aim of the investigation is to find out possible roles of adult neurogenesis in memory neural networks. As a first step, we numerically analysed the influence of randomly connecting external inputs on retrievals when the associative chaotic neural network retrievals the stored patterns with chaotic dynamics. As the result, the randomly connecting external inputs make less frequent retrieval of the stored patterns than the case without the inputs. However, the randomly connecting external inputs cause the network to switch the retrieval pattern more frequent than the case without the inputs.