Variable selection in finite mixture of regression models with an unknown number of components

被引：1

作者：

Lee, Kuo-Jung ^{[1
,2
]}

Feldkircher, Martin ^{[3
]}

Chen, Yi-Chi ^{[4
]}

机构：

[1] Natl Cheng Kung Univ, Dept Stat, Tainan, Taiwan

[2] Natl Cheng Kung Univ, Inst Data Sci, Tainan, Taiwan

[3] Vienna Sch Int Studies DA, Vienna, Austria

[4] Natl Cheng Kung Univ, Dept Econ, Tainan, Taiwan

来源：

COMPUTATIONAL STATISTICS & DATA ANALYSIS | 2021年 / 158卷

关键词：

Finite mixture of regression models; Bayesian variable selection; Unknown number of components; High-dimensional data; Financial crisis; BAYESIAN-ANALYSIS; REVERSIBLE JUMP; REGULARIZATION; LASSO; ORDER;

D O I：

10.1016/j.csda.2021.107180

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

A Bayesian framework for finite mixture models to deal with model selection and the selection of the number of mixture components simultaneously is presented. For that purpose, a feasible reversible jump Markov Chain Monte Carlo algorithm is proposed to model each component as a sparse regression model. This approach is made robust to outliers by using a prior that induces heavy tails and works well under multicollinearity and with high-dimensional data. Finally, the framework is applied to cross-sectional data investigating early warning indicators. The results reveal two distinct country groups for which estimated effects of vulnerability indicators vary considerably. (C) 2021 Elsevier B.V. All rights reserved.

引用

页数：19

共 50 条

[21] Finite mixture models do not reliably learn the number of components
Cai, Diana
Campbell, Trevor
Broderick, Tamara
[J]. INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 139, 2021, 139
[22] A new model selection procedure for finite mixture regression models
Yu, Conglian
Wang, Xiyang
[J]. COMMUNICATIONS IN STATISTICS-THEORY AND METHODS, 2020, 49 (18) : 4347 - 4366
[23] Variable selection in finite mixture of location and mean regression models using skew-normal distribution
Wang, Danlu
Zeng, Xin
Wu, Liucang
[J]. COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, 2023, 52 (12) : 5788 - 5810
[24] VARIABLE SELECTION IN REGRESSION-MODELS USING PRINCIPAL COMPONENTS
BONEH, S
MENDIETA, GR
[J]. COMMUNICATIONS IN STATISTICS-THEORY AND METHODS, 1994, 23 (01) : 197 - 213
[25] Selection of the number of components for finite mixtures of linear mixed models
Novais, Luisa
Faria, Susana
[J]. JOURNAL OF INTERDISCIPLINARY MATHEMATICS, 2021, 24 (08) : 2237 - 2268
[26] Physical mixture modeling with unknown number of components
Fischer, R
Dose, V
[J]. BAYESIAN INFERENCE AND MAXIMUM ENTROPY METHODS IN SCIENCE AND ENGINEERING, 2002, 617 : 143 - 154
[27] The Mixturegram: A Visualization Tool for Assessing the Number of Components in Finite Mixture Models
Young, Derek S.
Ke, Chenlu
Zeng, Xiaoxue
[J]. JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, 2018, 27 (03) : 564 - 575
[28] An Overview of the New Feature Selection Methods in Finite Mixture of Regression Models
Khalili, Abbas
[J]. JIRSS-JOURNAL OF THE IRANIAN STATISTICAL SOCIETY, 2011, 10 (02): : 201 - 235
[29] Bayesian model selection for mixtures of structural equation models with an unknown number of components
Lee, SY
Song, XY
[J]. BRITISH JOURNAL OF MATHEMATICAL & STATISTICAL PSYCHOLOGY, 2003, 56 : 145 - 165
[30] Bayesian analysis of mixture models with an unknown number of components - An alternative to reversible jump methods
Stephens, M
[J]. ANNALS OF STATISTICS, 2000, 28 (01): : 40 - 74

← 1 2 3 4 5 →