Bayesian regularization in constructive neural networks

Abstract

In this paper, we study the incorporation of Bayesian regularization into constructive neural networks. The degree of regularization is automatically controlled in the Bayesian inference framework and hence does not require manual setting. Simulation shows that regularization, with input training using a full Bayesian approach, produces networks with better generalization performance and lower susceptibility to over-fitting as the network size increases. Regularization with input training under MacKay's evidence framework, however, does not produce significant improvement on the problems tested.