CHAOS ASSOCIATIVE MEMORY WITH A PERIODIC ACTIVATION FUNCTION

In this paper we shall propose a novel chaos neural network model appl ied to the chaotic autoassociation memory. The present artificial neur on model is properly characterized in terms of a time-dependent period ic activation function to involve a chaotic dynamics as well as the en ergy steepest descent strategy. It is elucidated that the present neur al network has a remarkable ability of the dynamic memory retrievals b eyond the conventional models with the nonmonotonous activation functi on as well as such a monotonous activation function as sigmoidal one. This advantage is found to result from the property of the analogue pe riodic mapping accompanied with a chaotic behaviour of the neurons. It is also concluded that the present analogue neuron model with the per iodicity control has an apparently large memory capacity in comparison with the previously proposed association models.