Partially annealed neural networks
- 7 July 1994
- journal article
- Published by IOP Publishing in Journal of Physics A: General Physics
- Vol. 27 (13) , 4401-4411
- https://doi.org/10.1088/0305-4470/27/13/015
Abstract
We consider the Hopfield model of neural networks in which the patterns, as well as the spins, are dynamical variables. The characteristic time scales of the dynamics of the spins and the patterns are assumed to be widely separated such that the spins completely equilibrate at the time scale at which the elementary changes in the patterns take place. We study the situation in which each type of variable thermalizes at different temperatures, respectively, T and T'. In this case, such a system is described in terms of the traditional replica formalism in which the number of replicas n=T/T' is still the finite parameter. The complete phase diagram of the model in the space of the parameters T, alpha and n is obtained. If the parameter n is negative, the model is argued to present some similarities with the unlearning training algorithm. In this case a substantial increase in size of the retrieval phase in the plane (T, alpha ) is found.Keywords
This publication has 9 references indexed in Scilit:
- Statistical mechanics of neural networks near saturationPublished by Elsevier ,2004
- Partial annealing and overfrustration in disordered systemsJournal of Physics A: General Physics, 1994
- Coupled dynamics of fast spins and slow interactions in neural networks and spin systemsJournal of Physics A: General Physics, 1993
- Increasing the efficiency of a neural network through unlearningPhysica A: Statistical Mechanics and its Applications, 1990
- Associative recall of memory without errorsPhysical Review A, 1987
- "Unlearning" increases the storage capacity of content addressable memoriesBiophysical Journal, 1987
- Information storage and retrieval in spin-glass like neural networksJournal de Physique Lettres, 1985
- Neural networks and physical systems with emergent collective computational abilities.Proceedings of the National Academy of Sciences, 1982
- Solvable Model of a Spin-GlassPhysical Review Letters, 1975