On permutation symmetries of hopfield model neural network

Citation
Jy. Dong et al., On permutation symmetries of hopfield model neural network, DISCR D N S, 6(2), 2001, pp. 129-136
Citations number
14
Language
INGLESE
art.tipo
Article
Categorie Soggetti
Multidisciplinary
Journal title
DISCRETE DYNAMICS IN NATURE AND SOCIETY
ISSN journal
1026-0226 → ACNP
Volume
6
Issue
2
Year of publication
2001
Pages
129 - 136
Database
ISI
SICI code
1026-0226(2001)6:2<129:OPSOHM>2.0.ZU;2-9
Abstract
Discrete Hopfield neural network (DHNN) is studied by performing permutatio n operations on the synaptic weight matrix. The storable patterns set store d with Hebbian learning algorithm in a network without losing memories is s tudied, and a condition which makes sure all the patterns of the storable p atterns set have a same basin size of attraction is proposed. Then, the per mutation symmetries of the network are studied associating with the stored patterns set. A construction of the storable patterns set satisfying that c ondition is achieved by consideration of their invariance under a point gro up.