Influence of interconnection weight discretization and noise in an optoelectronic neural network
- 1 September 1989
- journal article
- Published by Optica Publishing Group in Optics Letters
- Vol. 14 (17) , 928-930
- https://doi.org/10.1364/ol.14.000928
Abstract
The effect of discretizing interconnection weight strengths in an optoelectronic learning neural network based on the backpropagation algorithm is investigated. We discuss how discretization arises in such an implementation. Using computer simulations we find that learning performance, as tested on the two-input XOR problem, is poor but that the addition of noise to the system results in substantial improvement.Keywords
This publication has 9 references indexed in Scilit:
- Optoelectronic analogs of self-programming neural nets: architecture and methodologies for implementing fast stochastic learning by simulated annealingApplied Optics, 1987
- Architectures for optoelectronic analogs of self-organizing neural networksOptics Letters, 1987
- One-Dimensional Silicon/PLZT Spatial Light ModulatorsOptical Engineering, 1987
- Two-Dimensional Magneto-Optic Spatial Light Modulator For Signal ProcessingOptical Engineering, 1983
- Optical matrix–matrix multiplier based on outer product decompositionApplied Optics, 1982
- Neural networks and physical systems with emergent collective computational abilities.Proceedings of the National Academy of Sciences, 1982
- Submicrosecond bistable electro-optic switching in liquid crystalsApplied Physics Letters, 1980
- Fully parallel, high-speed incoherent optical method for performing discrete Fourier transformsOptics Letters, 1978