Affordable Access

Publisher Website

Noise-induced order in the randomly asymmetric Hopfield model

Authors
Journal
Neural Networks
0893-6080
Publisher
Elsevier
Publication Date
Volume
8
Issue
2
Identifiers
DOI: 10.1016/0893-6080(94)00068-w
Keywords
  • Mathematical And Computational Analysis

Abstract

Abstract The generalized Hopfield model with randomly asymmetric bonds is studied. It is assumed that the synaptic connections { J ij } and the external inputs { U i } contain the fast noise and the frozen noise. By averaging over this randomness, we get the equilibrium distribution of the mean activity determined by the four parameters: the means and the variances of { U i } and { J ij }. This distribution has the term dynamic temperature, which is influenced by the mean activity through the variance of { J ij }. We analyze the phase transitions of the system. The result of analysis shows that increasing of the randomness can induce the transition from a monostable state to a bistable state. It differs from the ordinary phase transition where static temperature exists alone.

There are no comments yet on this publication. Be the first to share your thoughts.