Title :
Discrete dynamic neural memories: training and performance
Author :
Hassoun, Mohamad H.
Author_Institution :
Dept. of Electr. & Comput. Eng., Wayne State Univ., Detroit, MI, USA
Abstract :
A discrete dynamic associative neural memory (DAM) has a regular layered structure of simple neural processing units with feedback that is ideal for optimal implementations. DAM is reviewed for both autoassociative and heteroassociative processing. Several optimal DAM recording/learning algorithms and learning strategies are reviewed. Important DAM dynamics and performance characteristics are discussed and compared for various DAM synthesis techniques
Keywords :
content-addressable storage; learning systems; memory architecture; neural nets; DAM; autoassociative processing; discrete dynamic associative neural memory; feedback; heteroassociative processing; learning strategies; neural processing units; recording/learning algorithms; regular layered structure; synthesis techniques; Associative memory; Capacity planning; Convergence; Neurons; Nonlinear optical devices; Nonlinear optics; Optical arrays; Optical interconnections; Optical recording; Pain;
Conference_Titel :
Circuits and Systems, 1989., IEEE International Symposium on
Conference_Location :
Portland, OR
DOI :
10.1109/ISCAS.1989.100392
