Skip to content

Training a perceptron by a bit sequence: Storage capacity

M. Schroeder, W. Kinzel, I. Kanter

cond-matarXiv:cond-mat/9607040

Abstract

A perceptron is trained by a random bit sequence. In comparison to the corresponding classification problem, the storage capacity decreases to alphac=1.70 0.02 due to correlations between input and output bits. The numerical results are supported by a signal to noise analysis of Hebbian weights.

Create a lesson