Null Distribution Of The Likelihood Ratio Statistic For Feed-Forward Neural Networks
Open Access
- 1 November 2002
- journal article
- Published by Scilight Press Pty Ltd in Journal of Modern Applied Statistical Methods
- Vol. 1 (2) , 333-342
Abstract
Despite recent publications exploring model complexity with modern regression methods, their dimensionality is rarely quantified in practice and the distributions of related test statistics are not well characterized. Through a simulation study, we describe the null distribution of the likelihood ratio statistic for several different feed-forward neural network models.Keywords
This publication has 10 references indexed in Scilit:
- Measuring a Binary Response's Range of Influence in Logistic RegressionThe American Statistician, 2002
- Bayesian inference in neural networksBiometrika, 2001
- Counting degrees of freedom in hierarchical and other richly-parameterised modelsBiometrika, 2001
- On Measuring and Correcting the Effects of Data Mining and Model SelectionJournal of the American Statistical Association, 1998
- Modern Applied Statistics with S-PLUS1997
- Feed-forward Neural Networks1996
- Neural Network Studies. 2. Variable SelectionJournal of Chemical Information and Computer Sciences, 1996
- Maximum Likelihood Neural Network Prediction ModelsBiometrical Journal, 1995
- Unbiased estimate of generalization error and model selection in neural networkNeural Networks, 1995
- Statistical Theory of Learning Curves under Entropic Loss CriterionNeural Computation, 1993