Pattern Association

Associative memory neural nets store pattern associations, which are input-output vector pairs. They can be autoassociative (input equals output) or heteroassociative (input differs from output). Patterns are converted to binary or bipolar representations for computation