File(s) under permanent embargo
Order selection in finite mixture models: Complete or observed likelihood information criteria?
journal contribution
posted on 2023-05-19, 08:54 authored by Hui, FKC, Warton, DI, Scott FosterScott FosterChoosing the number of components in a finite mixture model is a challenging task. In this article, we study the behaviour of information criteria for selecting the mixture order, based on either the observed likelihood or the complete likelihood including component labels. We propose a new observed likelihood criterion called aicmix, which is shown to be order consistent. We further show that when there is a nontrivial level of classification uncertainty in the true model, complete likelihood criteria asymptotically underestimate the true number of components. A simulation study illustrates the potentially poor finite-sample performance of complete likelihood criteria, while aicmix and the Bayesian information criterion perform strongly regardless of the level of classification uncertainty.
History
Publication title
BiometrikaVolume
102Pagination
724-730ISSN
0006-3444Department/School
Institute for Marine and Antarctic StudiesPublisher
Biometrika TrustPlace of publication
Univ College London Gower St-Biometrika Office, London, England, Wc1E 6BtRights statement
Copyright 2015 Biometrika TrustRepository Status
- Restricted