Abstract

An optical perceptronlike neural network employing a delta learning rule and consisting of input units and a single output unit is constructed. Photorefractive crystals are used as holographic media for the interconnections between the input and the output layers. The learning rate for interconnection weight changes is optimally determined by setting the exposure time of the hologram. A learning experiment verifies the learning schedule's prediction that a fast, stable convergence of the learning process without oscillatory action can be obtained.

© 1989 Optical Society of America

All-optical error-signal generation for backpropagation learning in optical multilayer neural networks

Hisao Yoshinaga, Ken-ichi Kitayama, and Tsutomu Hara
Opt. Lett. 14(4) 202-204 (1989)

All-optical neural network with inhibitory neurons

I. Shariv and A. A. Friesem
Opt. Lett. 14(10) 485-487 (1989)

Demonstration of an optical inhibitory neural network

Wataru Kawakami, Hisao Yoshinaga, and Ken-ichi Kitayama
Opt. Lett. 14(18) 984-986 (1989)

Influence of interconnection weight discretization and noise in an optoelectronic neural network

A. Von Lehmen, E. G. Paek, P. F. Liao, A. Marrakchi, and J. S. Patel
Opt. Lett. 14(17) 928-930 (1989)

Continuous-time optical neural network associative memory

L.-S. Lee, H. M. Stoll, and M. C. Tackitt
Opt. Lett. 14(3) 162-164 (1989)