Posterior model probabilities via path-based pairwise priors

被引：33

作者：

Berger, JO

Molina, G

机构：

[1] Duke Univ, Res Triangle Pk, NC 27709 USA

[2] Stat & Appl Math Sci Inst, Res Triangle Pk, NC 27709 USA

[3] Credit Suisse First Boston, Fixed Income Res, London E14 4QJ, England

来源：

STATISTICA NEERLANDICA | 2005年 / 59卷 / 01期

关键词：

pairwise model comparisons; model selection; inclusion probabilities; search in model space;

D O I：

10.1111/j.1467-9574.2005.00275.x

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：

We focus on Bayesian model selection for the variable selection problem in large model spaces. The challenge is to search the huge model space adequately, while accurately approximating model posterior probabilities for the visited models. The issue of choice of prior distributions for the visited models is also important.

引用

页码：3 / 15

页数：13

共 15 条

[1] Optimal predictive model selection [J].

Barbieri, MM ;

Berger, JO .

ANNALS OF STATISTICS, 2004, 32 (03) :870-897

[2] The intrinsic Bayes factor for model selection and prediction [J].

Berger, JO ;

Pericchi, LR .

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1996, 91 (433) :109-122

[3]

BERGER JO, 2001, MODEL SELECTION, P135

[4]

BREIMAN L, 1985, J AM STAT ASSOC, V80, P580, DOI 10.2307/2288473

[5]

CASELLA R, 2002, 2002026 U FLOR STAT

[6]

Chipman H., 2001, Model selection, P67, DOI DOI 10.1214/LNMS/1215540964.655

[7] Bayesian CART model search [J].

Chipman, HA ;

George, EI ;

McCulloch, RE .

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1998, 93 (443) :935-948

[8]

Clyde MA, 1999, BAYESIAN STATISTICS 6, P157

[9] On Bayesian model and variable selection using MCMC [J].

Dellaportas, P ;

Forster, JJ ;

Ntzoufras, I .

STATISTICS AND COMPUTING, 2002, 12 (01) :27-36

[10]

DEVOS AF, 1993, FAIR COMP REGRESSION

← 1 2 →