Itinerant memory dynamics and global bifurcations in chaotic neural networks.

Abstract

We have considered itinerant memory dynamics in a chaotic neural network composed of four chaotic neurons with synaptic connections determined by two orthogonal stored patterns as a simple example of a chaotic itinerant phenomenon in dynamical associative memory. We have analyzed a mechanism of generating the itinerant memory dynamics with respect to intersection of a pair of alpha branches of periodic points and collapse of a periodic in-phase attracting set. The intersection of invariant sets is numerically verified by a novel method proposed in this paper.