The sparse Blume-Emery-Griffiths model of associative memories

TL;DR

This paper analyzes the sparse Blume-Emery-Griffiths (BEG) associative memory model at zero temperature, providing bounds on its storage capacity and demonstrating its superior performance over other sparse neural network models.

Contribution

It offers new theoretical bounds on the storage capacity of the sparse BEG model and compares its performance favorably to other models.

Findings

BEG model has higher storage capacity than comparable models

Bounds on storage capacity are established for fixed-point retrieval

BEG model outperforms other sparse neural network models in simulations

Abstract

We analyze the Blume-Emery-Griffiths (BEG) associative memory with sparse patterns and at zero temperature. We give bounds on its storage capacity provided that we want the stored patterns to be fixed points of the retrieval dynamics. We compare our results to that of other models of sparse neural networks and show that the BEG model has a superior performance compared to them.