Statistical mechanics of a multilayered neural network

E. Barkai, D. Hansel, I. Kanter

Research output: Contribution to journalArticlepeer-review

74 Scopus citations

Abstract

Statistical mechanics is applied to estimate the maximal information capacity per synapse (c) of a multilayered feedforward neural network, functioning as a parity machine. For a large number of hidden units, K, the replica-symmetric solution overestimates dramatically the capacity,K2. However, a one-step replica-symmetry breaking gives lnK/ln2, which coincides with a theoretical upper bound. It is suggested that this asymptotic behavior is exact. Results for finite K are also discussed.

Original languageEnglish
Pages (from-to)2312-2315
Number of pages4
JournalPhysical Review Letters
Volume65
Issue number18
DOIs
StatePublished - 1990

Fingerprint

Dive into the research topics of 'Statistical mechanics of a multilayered neural network'. Together they form a unique fingerprint.

Cite this