3 May 1988 New Recording Algorithm For Hopfield Model Associative Memories
Author Affiliations +
Proceedings Volume 0882, Neural Network Models for Optical Computing; (1988); doi: 10.1117/12.944101
Event: 1988 Los Angeles Symposium: O-E/LASE '88, 1988, Los Angeles, CA, United States
Abstract
A new recording technique for Hopfield-type associative/content address-able memories is proposed. The new technique is based on the finite and exponentially convergent algorithm of Ho and Kashyap[1] for the solution of a system of linear inequalities. Associative neural memories recorded with the proposed algorithm are shown to be superior to those recorded with the Hopfield's outer product and Kohonen's generalized inverse techniques. High capacity, high convergence rates to stored memories, and low convergence rates to false and oscillatory states are characteristics of this new recording algorithm. The issue of stable false and oscillatory states is raised, and it is shown that such states have a direct Boolean logic relationship with the stored memories.
© (1988) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
M H Hassoun, A M. Youssef, "New Recording Algorithm For Hopfield Model Associative Memories", Proc. SPIE 0882, Neural Network Models for Optical Computing, (3 May 1988); doi: 10.1117/12.944101; https://doi.org/10.1117/12.944101
PROCEEDINGS
9 PAGES


SHARE
KEYWORDS
Content addressable memory

Binary data

Logic

Image processing

Evolutionary algorithms

Neural networks

Neurons

RELATED CONTENT

Neural Networks For Logic Operations
Proceedings of SPIE (October 25 1989)
Template learning in morphological neural nets
Proceedings of SPIE (July 01 1991)
An Optoelectronic Neural Network
Proceedings of SPIE (February 05 1990)
Modeling of neural net chips using image algebra
Proceedings of SPIE (November 01 1990)

Back to Top