UNLEARNING IN NEURAL NETWORKS WITH MANY-NEURON INTERACTIONS

Authors
BARROZO MCD PENNA TJP
Citation
Mcd. Barrozo et Tjp. Penna, UNLEARNING IN NEURAL NETWORKS WITH MANY-NEURON INTERACTIONS, International journal of modern physics C, 5(3), 1994, pp. 503-512
Citations number
17
Categorie Soggetti
Mathematical Method, Physical Science","Physycs, Mathematical","Computer Science Interdisciplinary Applications
Journal title
International journal of modern physics C → ACNP
ISSN journal
01291831
Volume
5
Issue
3
Year of publication
1994
Pages
503 - 512
Database
ISI
SICI code
0129-1831(1994)5:3<503:UINNWM>2.0.ZU;2-K
Abstract
We present a version of a multineuron interaction model that includes unlearning. Through extensive numerical simulations we can show that u nlearning increases the performance of the network with respect to ret rieval of stored informations by enlarging the basins of attraction. R elationships that allow us to estimate the optimal performance in the learning stage of this model were found.