Title :
Guaranteed recall of all training pairs for bidirectional associative memory
Author :
Wang, Yeou-Fang ; Cruz, Jose B., Jr. ; Mulligan, J.H., Jr.
Author_Institution :
Dept. of Electr. & Comput. Eng., California Univ., Irvine, CA, USA
fDate :
11/1/1991 12:00:00 AM
Abstract :
Necessary and sufficient conditions are derived for the weights of a generalized correlation matrix of a bidirectional associative memory (BAM) which guarantee the recall of all training pairs. A linear programming/multiple training (LP/MT) method that determines weights which satisfy the conditions when a solution is feasible is presented. The sequential multiple training (SMT) method is shown to yield integers for the weights, which are multiplicities of the training pairs. Computer simulation results, including capacity comparisons of BAM, LP/MT BAM, and SMT BAM, are presented
Keywords :
content-addressable storage; learning systems; linear programming; matrix algebra; neural nets; bidirectional associative memory; capacity comparisons; content addressable storage; generalized correlation matrix; linear programming; neural nets; sequential multiple training; training pairs; Application software; Associative memory; Computer simulation; Linear programming; Magnesium compounds; Manufacturing automation; Neural networks; Subcontracting; Sufficient conditions; Surface-mount technology;
Journal_Title :
Neural Networks, IEEE Transactions on
