A note on BIC in mixed-effects models

被引：82

作者：

Delattre, Maud ^{[1
,2
]}

Lavielle, Marc ^{[3
]}

Poursat, Marie-Anne ^{[4
]}

机构：

[1] AgroParisTech, UMR MIA 518, F-75005 Paris, France

[2] INRA, UMR MIA 518, F-75005 Paris, France

[3] Univ Paris 11, Inria Saclay Ile France, Lab Math UMR 8628, F-91405 Orsay, France

[4] Univ Paris 11, Lab Math UMR 8628, F-91405 Orsay, France

来源：

ELECTRONIC JOURNAL OF STATISTICS | 2014年 / 8卷

关键词：

Bayesian Information Criterion; BIC; mixed effects model; variable selection; CONDITIONAL AKAIKE INFORMATION; APPROXIMATION; CONVERGENCE;

D O I：

10.1214/14-EJS890

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

The Bayesian Information Criterion (BIC) is widely used for variables election in mixed effects models. However, its expression is unclear in typical situations of mixed effects models, where simple definition of the sample size is not meaningful. We derive an appropriate BIC expression that is consistent with the random effect structure of the mixed effects model. We illustrate the behavior of the proposed criterion through a simulation experiment and a case study and we recommend its use as an alternative to various existing BIC versions that are implemented in available software

引用

页码：456 / 475

页数：20

共 28 条

[1]

[Anonymous], 2006, J. de la Societe Francaise de Statistique

[2]

[Anonymous], 2009, MIXED EFFECTS MODELS

[3]

[Anonymous], 2008, SASSTAT 92 USERS GUI, P4337

[4]

Bondell H., 2010, BIOMETRICS, V66, P1069

[5] Nonlinear models for repeated measurement data: An overview and update [J].

Davidian, M ;

Giltinan, DM .

JOURNAL OF AGRICULTURAL BIOLOGICAL AND ENVIRONMENTAL STATISTICS, 2003, 8 (04) :387-419

[6] Maximum likelihood estimation in discrete mixed hidden Markov models using the SAEM algorithm [J].

Delattre, M. ;

Lavielle, M. .

COMPUTATIONAL STATISTICS & DATA ANALYSIS, 2012, 56 (06) :2073-2085

[7]

DELATTRE M, 2012, THESIS U PARIS SUD

[8]

Delyon B, 1999, ANN STAT, V27, P94

[9] Conditional Akaike information under generalized linear and proportional hazards mixed models [J].

Donohue, M. C. ;

Overholser, R. ;

Xu, R. ;

Vaida, F. .

BIOMETRIKA, 2011, 98 (03) :685-700

[10]

Jiang J., 2003, Sankhya, V65, P23

← 1 2 3 →