This article reviews current advances and developments in neural networks. This requires recalling some of the earlier work in the field. We emphasize Bayesian approaches and their benefits compared to more standard maximum likelihood treatments. Several representative experiments using varied...