A Novel Associative Memory System Based on Newton's Forward Interpolation

Abstract

A novel highorder associative memory system based on the Newton's forward Interpolation (NFIAMS) is preposed, which can carry out errorfree approximations to multivariable polynomial functions of arbitrary order. The theory, interpolation algorithm and traning rules of the NFIAMS are discussed in detail. The results of simulating examples indicate that it has more advantadges than CMACtype AMS, such as highprecision of learning, much smaller memory requirement without the datacollision problem and much less computational effort for training and faster convergence rates than that attainable with multilayer BP neural networks.