Self-control in Sparsely Coded Networks

TL;DR

This paper introduces a self-control mechanism in neural networks using a dynamic threshold that adapts to noise and activity, significantly enhancing performance in sparsely coded models.

Contribution

It presents a novel self-control mechanism with a time-dependent threshold that improves storage capacity and information content in neural networks.

Findings

Increased storage capacity in sparse networks

Enhanced basins of attraction

Improved mutual information content

Abstract

A complete self-control mechanism is proposed in the dynamics of neural networks through the introduction of a time-dependent threshold, determined in function of both the noise and the pattern activity in the network. Especially for sparsely coded models this mechanism is shown to considerably improve the storage capacity, the basins of attraction and the mutual information content of the network.