Title :
CMOS mean field learning
Author :
Schneider, Charles R. ; Card, H.C.
Author_Institution :
Dept. of Electr. & Comput. Eng., Manitoba Univ., Winnipeg, Man., Canada
Abstract :
Analogue CMOS circuits which implement contrastive Hebbian learning have been designed, fabricated, and tested. A 19 neuron, 342 synapse mean field network, or deterministic Boltzmann machine (4.16*2.76 mm2 in 1.2 mu m CMOS) has been constructed based on these synaptic circuits. Very small learning step size may be employed in this chip, which is advantageous for reliable mean field learning.
Keywords :
CMOS integrated circuits; analogue circuits; learning systems; neural nets; analogue CMOS circuits; contrastive Hebbian learning; deterministic Boltzmann machine; learning step size; mean field network; synaptic circuits;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19911060
