A Continuous-Time Asynchronous Boltzmann Machine.

Miyajima, Teruyuki; Agu, Masahiro; Yamanaka, Kazuo · Neural Netw · 1997

basic_science · Level V

Where this comes from

Abstract

We propose an asynchronous neural network model having the same structure as the binary Hopfield model. Each neuron operates with continuous time and randomly changes its state according only to its membrane potential. The proposed model settles in a steady-state fluctuation, in which the probability distribution of the global state is identical to that of the serial Boltzmann machine with the same synaptic weights. Copyright 1997 Elsevier Science Ltd.