线性模型下基于AIC准测的Bayes变量选择

J4 ›› 2010, Vol. 45 ›› Issue (6): 43-45.

线性模型下基于AIC准测的Bayes变量选择

王树云¹,宋云胜²

1.山东大学数学学院, 山东济南 250100; 2.山西大学数学科学学院, 山西太原 030006

收稿日期:2010-02-01 出版日期:2010-06-16 发布日期:2010-06-17
作者简介:王树云(1981-),男,博士研究生,研究方向为概率论与数理统计. Email: shuyun.2001@mail.sdu.edu.cn
基金资助:
国家自然科学基金资助项目（10671110）；国家重点基础研究发展规划项目计划(973计划)(2007CB814900)

Bayesian variable selection based on AIC criteria in linear models

WANG Shu-yun¹, SONG Yun-sheng²

1. School of Mathematics, Shandong University, Jinan 250100, Shandong, China;
2. School of Mathematics, Shanxi University, Taiyuan 030006, Shanxi, China

Received:2010-02-01 Online:2010-06-16 Published:2010-06-17

摘要/Abstract

摘要：

讨论了线性模型下Bayes变量选择问题。通过用AIC准则来修正经典的Bayes变量选择方法,构造修正后的子模型后验分布,并且通过仿真计算验证,修正后的后验分布可以提高变量选择精度。

关键词: Bayes变量选择;AIC准则;后验分布

Abstract:

The problem of Bayesian variable selection in linear models is studied. Different with classical method of Bayesian variable selection, AIC criteria is used to construct the posterior distribution of the subset model. Simulation studies are also provided to demonstrate the better performance of the proposed method.

Key words: Bayesian variable selection; AIC criteria; Posterior distribution

王树云1,宋云胜2. 线性模型下基于AIC准测的Bayes变量选择[J]. J4, 2010, 45(6): 43-45.

WANG Shu-yun1, SONG Yun-sheng2. Bayesian variable selection based on AIC criteria in linear models[J]. J4, 2010, 45(6): 43-45.

参考文献

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed