Summary
International Symposium on Nonlinear Theory and Its Applications
2015
Session Number:B4L-B
Session:
Number:B4L-B-3
Chaotic Associative Memory Dynamics of Chaotic Neural Network Model with Time Dependent System Parameter
Tatsuhito Tamamura, Jousuke Kuroiwa, Tomohiro Odaka, Izumi Suwa, Haruhiko Shirai,
pp.680-683
Publication Date:2015/12/1
Online ISSN:2188-5079
Summary:
In this paper, we investigate chaotic associative memory dynamics in a chaotic neural network model (referred as CNN hereafter) with a time-dependent system parameter. We have shown that an isolated chaotic neuron model with a time-dependent system parameter gives attractor coexistence behaviors depending on initial conditions. In this paper, we introduce a time dependent system parameter into Adachi & Aihara CNN with association recalling dynamics. Consequently, the system possesses two types of initial dependence, originating in (i) synaptic connections and (ii) a time-dependent system parameter. The purposes of this paper are (i) to show whether two types of initial dependence could coexist or not, and (ii) to investigate chaotic associative memory dynamics for various initial configurations. From computer experiments, two types of initial dependence can coexist and the system shows complex associative memory dynamics. In several parameter regions, the system reveals different chaotic associative memory dynamics depending on different initial configurations of memory patterns, that is, the difference originates in synaptic connections. On the other hands, for the other initial configurations which are slightly different from memory patterns, the system shows different periodic orbits. The difference between chaotic and periodic originates in a time-dependent system parameter.