MULTIVALLEY STRUCTURE OF ATTRACTOR NEURAL NETWORKS

Authors
NETO CR FONTANARI JF
Citation
Cr. Neto et Jf. Fontanari, MULTIVALLEY STRUCTURE OF ATTRACTOR NEURAL NETWORKS, Journal of physics. A, mathematical and general, 30(22), 1997, pp. 7945-7951
Citations number
20
Categorie Soggetti
Physics,"Physycs, Mathematical
Journal title
Journal of physics. A, mathematical and general → ACNP
ISSN journal
03054470
Volume
30
Issue
22
Year of publication
1997
Pages
7945 - 7951
Database
ISI
SICI code
0305-4470(1997)30:22<7945:MSOANN>2.0.ZU;2-Z
Abstract
We numerically investigate the structure of the basins of attraction a nd the nature of the spurious attractors of the pseudo-inverse and the optimal weights attractor neural networks. We show that the number of attractors in the optimal weights model increases as the margin param eter kappa Increases, and that the basins of attraction of the stored patterns are not significantly enlarged by kappa. Moreover, the number of attractors is smaller than in the pseudo-inverse model.