Asymmetric Boltzmann machines.

B Apolloni, A Bertoni, P Campadelli, D de Falco
Author Information
  1. B Apolloni: Dipartimento di Scienze dell' Informazione, Università di Milano, Italy.

Abstract

We study asymmetric stochastic networks from two points of view: combinatorial optimization and learning algorithms based on relative entropy minimization. We show that there are non trivial classes of asymmetric networks which admit a Lyapunov function L under deterministic parallel evolution and prove that the stochastic augmentation of such networks amounts to a stochastic search for global minima of L. The problem of minimizing L for a totally antisymmetric parallel network is shown to be associated to an NP-complete decision problem. The study of entropic learning for general asymmetric networks, performed in the non equilibrium, time dependent formalism, leads to a Hebbian rule based on time averages over the past history of the system. The general algorithm for asymmetric networks is tested on a feed-forward architecture.

References

J Neurophysiol. 1946 May;9:191-204 [PMID: 21028162]
Proc Natl Acad Sci U S A. 1982 Apr;79(8):2554-8 [PMID: 6953413]

MeSH Term

Algorithms
Animals
Learning
Mathematics
Models, Neurological
Nerve Net
Software
Stochastic Processes
Synapses

Word Cloud

Similar Articles

Cited By