NEW LEARNING-STRATEGIES FROM THE MICROSCOPIC LEVEL OF AN ARTIFICIAL NEURAL-NETWORK

Authors
JOYA G FRIAS JJ MARIN MM SANDOVAL F
Citation
G. Joya et al., NEW LEARNING-STRATEGIES FROM THE MICROSCOPIC LEVEL OF AN ARTIFICIAL NEURAL-NETWORK, Electronics Letters, 29(20), 1993, pp. 1775-1777
Citations number
2
Categorie Soggetti
Engineering, Eletrical & Electronic
Journal title
Electronics Letters → ACNP
ISSN journal
00135194
Volume
29
Issue
20
Year of publication
1993
Pages
1775 - 1777
Database
ISI
SICI code
0013-5194(1993)29:20<1775:NLFTML>2.0.ZU;2-4
Abstract
New learning strategies deduced from a microscopic level study of the weights space of an artificial neural network (ANN) are presented. The se strategies improve the efficiency of learning algorithms, reducing the amount of patterns to be presented to the ANN.