The models of associative memory pioneered by Kosko  is a by-product of his research in fuzzy systems. Associative memory to some eXtent has resemblance with the human brain, where from one set of given input output pattern, the brain can determine analogously the output (input) pattern when the input (output) pattern is known. In Kosko's model, he assumed the fuzzy membership distribution of the output patterns for a given input pattern and
designed the weights (W) of a single-layered feed-forward neural net by taking the product of input feeder (I) and the transpose of the output feeder (O) :
Was this article helpful?