Hopfield learning rule with high capacity storage of time-correlated patterns

A Storkey*, R Valabregue

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

A new local and incremental learning rule is examined for its ability to store patterns from a time series in an attractor neural network. This learning rule has a higher capacity than the Hebb rule, and suffers significantly less capacity loss as the correlation between patterns increases.

Original languageEnglish
Pages (from-to)1803-1804
Number of pages2
JournalElectronics Letters
Volume33
Issue number21
DOIs
Publication statusPublished - 9 Oct 1997

Keywords

  • Hopfield neural networks
  • MEMORY

Cite this