Using Bayesian Methods for Avoiding Overfitting and for Ranking Networks in Multilayer Perceptrons Learning
This work is an experimental attempt to determine whether the Bayesian paradigm could improve Multilayer Perceptrons (MLPs) learning methods. In particular, we experiment here the paradigm developed by D. MacKay. The paper points out the main or critical points of MacKay's work and introduces very practical points of Bayesian MLPs, having in mind future applications. Bayesian MLPs are used on 3 public classification databases and compared to other methods.
Bibliographic Reference: Article: Neurocolt T.R. Series (1995)
Record Number: 199610123 / Last updated on: 1996-03-01
Original language: en
Available languages: en