Testing the order of a model

DOI10.1214/009053606000000344MaRDI QIDQ2500450zbMATH OpenOpenAlexFDO

Authors Antoine Chambaz

Publication date 24 August 2006

Published in The Annals of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/math/0607806

zbMATH Keywords

abrupt changesempirical processeserror exponentshypothesis testinglarge deviationsmixturesmodel selectionmoderate deviationsorder estimation

Mathematics Subject Classification ID

Applications of functional analysis in probability theory and statistics (46N30) Large deviations (60F10) Random measures (60G57) Statistical decision theory (62C99) Asymptotic properties of parametric tests (62F05) Asymptotic properties of parametric estimators (62F12)

Abstract: This paper deals with order identification for nested models in the i.i.d. framework. We study the asymptotic efficiency of two generalized likelihood ratio tests of the order. They are based on two estimators which are proved to be strongly consistent. A version of Stein's lemma yields an optimal underestimation error exponent. The lemma also implies that the overestimation error exponent is necessarily trivial. Our tests admit nontrivial underestimation error exponents. The optimal underestimation error exponent is achieved in some situations. The overestimation error can decay exponentially with respect to a positive power of the number of observations. These results are proved under mild assumptions by relating the underestimation (resp. overestimation) error to large (resp. moderate) deviations of the log-likelihood process. In particular, it is not necessary that the classical Cram'{e}r condition be satisfied; namely, the $log$-densities are not required to admit every exponential moment. Three benchmark examples with specific difficulties (location mixture of normal distributions, abrupt changes and various regressions) are detailed so as to illustrate the generality of our results.

Recommendations

• Error exponents for AR order testing
• Asymptotic Probabilities of Over-estimating and Under-estimating the Order of a Model in General Regular Families
• The optimal error exponent for Markov order estimation
• ORDER IDENTIFICATION IN MISSPECIFIED AUTOREGRESSIVE TIME SERIES MODELS
• Bounds for Bayesian order identification with application to mixtures

Cites work

• A new look at the statistical model identification
• An extension of Sanov's theorem: Application to the Gibbs conditioning principle
• Asymptotic Probabilities of Over-estimating and Under-estimating the Order of a Model in General Regular Families
• Asymptotic Statistics
• Consistent estimation of a mixing distribution
• Consistent estimation of mixture complexity.
• Consistent estimation of the order of mixture models.
• Convex Analysis
• Cramér's condition and Sanov's theorem
• Error exponents for AR order testing
• Estimating the dimension of a model
• scientific article; zbMATH DE number 3908323 (Why is no real title available?)
• scientific article; zbMATH DE number 3954108 (Why is no real title available?)
• scientific article; zbMATH DE number 3969875 (Why is no real title available?)
• scientific article; zbMATH DE number 4106081 (Why is no real title available?)
• scientific article; zbMATH DE number 469332 (Why is no real title available?)
• scientific article; zbMATH DE number 1153603 (Why is no real title available?)
• scientific article; zbMATH DE number 1158743 (Why is no real title available?)
• scientific article; zbMATH DE number 2062987 (Why is no real title available?)
• scientific article; zbMATH DE number 4117660 (Why is no real title available?)
• scientific article; zbMATH DE number 3335636 (Why is no real title available?)
• scientific article; zbMATH DE number 3336465 (Why is no real title available?)
• scientific article; zbMATH DE number 3406955 (Why is no real title available?)
• I-divergence geometry of probability distributions and minimization problems
• Invariance principles for sums of Banach space valued random elements and empirical processes
• Large deviations, moderate deviations and LIL for empirical processes
• Large-Sample Theory: Parametric Case
• Large-scale typicality of Markov sample paths and consistency of MDL order estimators
• Likelihood ratio inequalities with applications to various mixtures
• Modeling by shortest data description
• On estimating the number of constituents of a finite mixture of continuous distributions
• On the underfitting and overfitting sets of models chosen by order selection criteria.
• Optimal error exponents in hidden markov models order estimation
• Risk bounds for model selection via penalization
• Some applications of concentration inequalities to statistics
• Some Comments on C P
• Statistical analysis of finite mixture distributions
• Testing the order of a model using locally conic parametrization: Population mixtures and stationary ARMA processes
• The consistency of the BIC Markov order estimator.
• The estimation of the order of a mixture model
• The optimal error exponent for Markov order estimation

Cited in

(11)

• A minimum description length approach to hidden Markov models with Poisson and Gaussian emissions. Application to order identification
• Measurement by subjective estimation: Testing for separable representations
• Bounds for Bayesian order identification with application to mixtures
• The loss rank criterion for variable selection in linear regression analysis
• Error exponents for AR order testing
• Asymptotic Probabilities of Over-estimating and Under-estimating the Order of a Model in General Regular Families
• Test of Significance in order selection
• The estimation of the model order in exponential families
• The likelihood ratio test for general mixture models with or without structural parameter
• Covariate selection in mixture models with the censored response variable
• Model selection with the loss rank principle

This page was built for publication: Testing the order of a model

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2500450)