A novel generalized flip-flop for memory association and maximization in artificial neural network

A circuit structure which performs perfect memory association for the case where the stored memory patterns are any set of orthogonal vectors in an n -dimensional space is proposed. This circuit is based on a new interpretation of the classical flip-flop which is the basis of the majority of current circuit implementations of artificial neural networks. The result shows that for the simplified memory association problem considered, which so far has not been satisfactorily solved, a natural generalization of the flip-flop to higher dimensions provides a simple and elegant solution