Reduced complexity RBF classifiers with support vector centres and dynamic decay adjustment

Abstract

A fast and simple method is proposed to build low complexity radial basis function (RBF) classifiers. It is based on the approximation of the decision rule of a support vector machine by an RBF network, and integrates the dynamic decay adjustment algorithm with selective pruning and standard least squares techniques. Experimental results on several benchmark data sets, concerning both binary and multi-class problems, show the effectiveness of the proposed method.