A CHAOS ASSOCIATION MODEL WITH A TIME-DEPENDENT PERIODIC ACTIVATION FUNCTION

In this paper, an autoassociation model is proposed that uses a piecew ise linear periodic mapping with parameter control. In practice, the a ctivation function in the present model is controlled through a nonmon otonic periodic mapping between monotonic periodic mappings, so that i t realizes high memory capacity without falling into a false memory th at is a quasi-stable state of the network. In the present model, by us ing chaotic dynamics based on the periodic mapping, a memory search mo del using chaos can be constructed, Further, by providing information on the input, autoassociation can be realized. From the results of a c omputer simulation, it is found that the present model can realize per fect association with a memory rate of L/N similar to 0.5 even in the search mode. Further, in asynchronous operation of the network, improv ement of the association characteristics and enhancement of the memory capacity (with a memory rate of L/N similar to 0.56 in the search mod e) are confirmed, so that the usefulness of chaos in asynchronous oper ation is demonstrated, (C) 1998 Scripta Technica.