Statistical Analysis of an Autoassociative Memory Network

1 September 1994

journal article
Published by MIT Press in Neural Computation

Vol. 6 (5) , 837-841
https://doi.org/10.1162/neco.1994.6.5.837

Abstract

A statistical method for analyzing the dynamic behavior of a synchronous autoassociative memory network is presented based on an extension of Amari-Maginu theory. Through computer simulations it is shown that when the memorized pattern ratio is small (r < 0.15) or large (r > 0.2), the derived theory shows good agreement.

Keywords

ASYNCHRONOUS AUTOASSOCIATIVE MEMORY NETWORK
DYNAMIC BEHAVIOR
MEMORIZED PATTERN RATIO
AMARI-MAGINU THEORY
STATISTICAL ANALYSIS
THEORY SHOWSGOOD AGREEMENT
STATISTICAL METHOD
COMPUTER SIMULATION

This publication has 2 references indexed in Scilit:

Statistical neurodynamics of associative memory
Neural Networks, 1988
Storing Infinite Numbers of Patterns in a Spin-Glass Model of Neural Networks
Physical Review Letters, 1985