Basis function models of the CMAC network

An interpretation of the Cerebellar Model Articulation Controller (CMAC) ne twork as a member of the General Memory Neural Network (GMNN) architecture is presented. The usefulness of this approach stems from the fact that, wit hin the GMNN formalism, CMAC can be treated as a particular form of a basis function network, where the basis function is inherently dependent on the type of input quantization present in the network mapping. Furthermore, con sidering the relative regularity of input-space quantization performed by C MAC, we are able to derive an expected (or average) form of the basis funct ion characteristic of this network. Using this basis form, it is possible t o create basis-functions models of CMAC mapping, as well as to gain More in sight into its performance. The developments are supported by numerical sim ulations. (C) 1999 Elsevier Science Ltd. All rights reserved.