Normalized ML and the MDL principle for variable selection in linear regression

MaRDI QIDQ5455538zbMATH OpenFDO

Authors Erkki P. Liski

Publication date 3 April 2008

zbMATH Keywords

stochastic complexityparametric complexityadaptive selectionfixed penalty criteria

Mathematics Subject Classification ID

Statistical aspects of information-theoretic topics (62B10) Linear regression; mixed models (62J05) Parametric inference (62F99)

Recommendations

• Variable selection in linear regression: several approaches based on normalized maximum likelihood
• Model selection by normalized maximum likelihood
• Model selection with informative normalized maximum likelihood: data prior and model prior
• Stochastic complexity and the mdl principle
• On the normalized maximum likelihood and Bayesian decision theory

Cited in

(5)

• The decomposed normalized maximum likelihood code-length criterion for selecting hierarchical latent variable models
• Variable selection in linear regression: several approaches based on normalized maximum likelihood
• High-dimensional penalty selection via minimum description length principle
• Normalized maximum likelihood models for logit regression
• Classification and feature gene selection using the normalized maximum likelihood model for discrete regression

This page was built for publication: Normalized ML and the MDL principle for variable selection in linear regression

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5455538)