A LEARNING MULTIPLE-VALUED LOGIC NETWORK - ALGEBRA, ALGORITHM, AND APPLICATIONS

We propose a multiple-valued logic (MVL) network with functional compl eteness and develop its learning capability. The MVL network consists of layered arithmetic piecewise linear processors. Since the arithmeti c operations of the network are basically a wired-sum and a piecewise linear operation, their implementations should be rather simple and st raightforward. Furthermore, the MVL network can be trained by the trad itional backpropagation algorithm directly. The algorithm trains the n etworks using examples and appears to be available for most MVL proble ms of interest.