Sufficient dimension reduction via bayesian mixture modeling.

TitleSufficient dimension reduction via bayesian mixture modeling.
Publication TypeJournal Article
Year of Publication2011
AuthorsReich, Brian J., Howard D. Bondell, and Lexin Li
JournalBiometrics
Volume67
Issue3
Pagination886-95
Date Published2011 Sep
ISSN1541-0420
KeywordsBayes Theorem, Data Interpretation, Statistical, HIV, Humans, Models, Statistical
Abstract

Dimension reduction is central to an analysis of data with many predictors. Sufficient dimension reduction aims to identify the smallest possible number of linear combinations of the predictors, called the sufficient predictors, that retain all of the information in the predictors about the response distribution. In this article, we propose a Bayesian solution for sufficient dimension reduction. We directly model the response density in terms of the sufficient predictors using a finite mixture model. This approach is computationally efficient and offers a unified framework to handle categorical predictors, missing predictors, and Bayesian variable selection. We illustrate the method using both a simulation study and an analysis of an HIV data set.

DOI10.1111/j.1541-0420.2010.01501.x
Alternate JournalBiometrics
Original PublicationSufficient dimension reduction via bayesian mixture modeling.
PubMed ID21039398
PubMed Central IDPMC3117934
Grant ListP01 CA142538 / CA / NCI NIH HHS / United States
R01 ES014843 / ES / NIEHS NIH HHS / United States
R01 ES014843-01A2 / ES / NIEHS NIH HHS / United States
Project: