N Brunel et al 1992 J. Phys. A: Math. Gen. 25 5017 doi:10.1088/0305-4470/25/19/015
N Brunel, J -P Nadal and G Toulouse
Show affiliationsThe authors study the information storage capacity of a simple perceptron in the error regime. For random unbiased patterns the geometrical analysis gives a logarithmic dependence for the information content in the asymptotic limit. In this case, the statistical physics approach, when used at the simplest level of replica theory, does not give satisfactory results. However for perceptrons with finite stability, the information content can be simply calculated with statistical physics methods in a region above the critical storage level, for biased as well as for unbiased patterns.
07.05.Mh Neural networks, fuzzy logic, artificial intelligence
Issue 19 (7 October 1992)
N Brunel et al 1992 J. Phys. A: Math. Gen. 25 5017
A W Sandvik 1992 J. Phys. A: Math. Gen. 25 3667
Y S Yang and C J Thompson 1991 J. Phys. A: Math. Gen. 24 L279
E B Lin 1991 J. Phys. A: Math. Gen. 24 L1045
M Wilkinson et al 1991 J. Phys. A: Math. Gen. 24 175
V Pasquier 1987 J. Phys. A: Math. Gen. 20 L221
K Svozil 1987 J. Phys. A: Math. Gen. 20 3861
J L Cardy 1986 J. Phys. A: Math. Gen. 19 L1093
F Family and T Vicsek 1985 J. Phys. A: Math. Gen. 18 L75
U Bruzzo and R Cianci 1985 J. Phys. A: Math. Gen. 18 417