Chaotic Complex-Valued Multidirectional Associative Memory with Variable Scaling Factor

In this paper, we propose a Chaotic Complex-valued Multidirectional Associative Memory (CCMAM) with variable scale factor which can realize one-to-many associations of M-tuple multi-valued patterns. The proposed model is based on the Multidirectional Associative Memory, and is composed of complex-valued neurons and chaotic complex-valued neurons. In the proposed model, associations of multivalued patterns are realized by using complex-valued neurons, and one-to-many associations are realized by using chaotic complex-valued neurons. Moreover, in the proposed model, the appropriate parameters of chaotic complex-valued neurons can be determined easily than in the original Chaotic Complex-valued Multidirectional Associative Memory. We carried out a series of computer experiments and confirmed that the proposed model has superior one-to-many association ability than that of the conventional model.