A Comparative Study of Sparse Associative Memories

Gripon V., Heusel J., Löwe M., Vermet F.

Research article (journal) | Peer reviewed

Abstract

We study various models of associative memories with sparse information, i.e. a pattern to be stored is a random string of 0s and 1s with about log N 1s, only. We compare different synaptic weights, architectures and retrieval mechanisms to shed light on the influence of the various parameters on the storage capacity.

Details about the publication

Journal: Journal of Statistical Physics

Volume: 164

Issue: 1

Page range: 105-129

Status: Published

Release year: 2016

Language in which the publication is written: English

DOI: 10.1007/s10955-016-1530-z

Link to the full text: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84968547414&origin=inward

Keywords: Associative memory; Exponential inequalities; Neural networks; Sparse patterns; Storage capacity

Authors from the University of Münster

Löwe, Matthias

Professur für Mathematische Stochastik (Prof. Löwe)