TY - JOUR

T1 - Statistical mechanics of a multilayered neural network

AU - Barkai, E.

AU - Hansel, D.

AU - Kanter, I.

PY - 1990

Y1 - 1990

N2 - 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.

AB - 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.

UR - http://www.scopus.com/inward/record.url?scp=0000745533&partnerID=8YFLogxK

U2 - 10.1103/physrevlett.65.2312

DO - 10.1103/physrevlett.65.2312

M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???

AN - SCOPUS:0000745533

SN - 0031-9007

VL - 65

SP - 2312

EP - 2315

JO - Physical Review Letters

JF - Physical Review Letters

IS - 18

ER -