A digital CMOS fully connected neural network with in-circuit learning capability and automatic identification of spurious attractors

Abstract : Summary form only given. An electronic implementation of a completely connected feedback network, containing 64 neurons, is considered. The technology is fully digital CMOS, with binary neurons and 9-bit-wide signed synaptic coefficients. The architecture trades off connectivity versus speed by implementing a linear systolic loop, in which each neuron locally stores its own synaptic coefficients. The authors have first implemented internal learning capabilities. They used the Widrow-Hoff rule, which converges towards the projection rule by iteration, thus allowing partial correlation between prototypes and a higher capacity compared to the Hebb rule. They have also implemented an internal mechanism for detecting relaxations on spurious states. The combination of these two properties gives the network a rather high degree of autonomy, making unnecessary the use of an external computer for tasks other than just writing or reading data and asserting simple control signals.
Type de document :
Communication dans un congrès
Neural Networks, 1991., IJCNN-91-Seattle International Joint Conference on, Jul 1991, Seattle, United States. 2, pp.923, 1991, 〈10.1109/IJCNN.1991.155576〉
Liste complète des métadonnées

https://hal.inria.fr/hal-00700101
Contributeur : Jean-Dominique Gascuel <>
Soumis le : mardi 22 mai 2012 - 12:30:59
Dernière modification le : jeudi 12 avril 2018 - 01:45:36

Identifiants

Collections

Citation

J.D. Gascuel, M. Weinfeld. A digital CMOS fully connected neural network with in-circuit learning capability and automatic identification of spurious attractors. Neural Networks, 1991., IJCNN-91-Seattle International Joint Conference on, Jul 1991, Seattle, United States. 2, pp.923, 1991, 〈10.1109/IJCNN.1991.155576〉. 〈hal-00700101〉

Partager

Métriques

Consultations de la notice

86