Two-level stochastic search variable selection in GLMs with missing predictors
Two-level stochastic search variable selection in GLMs with missing predictors
Stochastic search variable selection (SSVS) algorithms provide an appealing and widely used approach for searching for good subsets of predictors while simultaneously estimating posterior model probabilities and model-averaged predictive distributions. This article proposes a two-level generalization of SSVS to account for missing predictors while accommodating uncertainty in the relationships between these predictors. Bayesian approaches for allowing predictors that are missing at random require a model on the joint distribution of the predictors. We show that predictive performance can be improved by allowing uncertainty in the specification of predictor relationships in this model. The methods are illustrated through simulation studies and analysis of an epidemiologic data set.
33
Mitra, Robin
2b944cd7-5be8-4dd1-ab44-f8ada9a33405
Dunson, David
560f17c5-3851-4163-a8e4-d3b12ad9650c
January 2010
Mitra, Robin
2b944cd7-5be8-4dd1-ab44-f8ada9a33405
Dunson, David
560f17c5-3851-4163-a8e4-d3b12ad9650c
Mitra, Robin and Dunson, David
(2010)
Two-level stochastic search variable selection in GLMs with missing predictors.
International Journal of Biostatistics, 6 (1), .
(doi:10.2202/1557-4679.1173).
Abstract
Stochastic search variable selection (SSVS) algorithms provide an appealing and widely used approach for searching for good subsets of predictors while simultaneously estimating posterior model probabilities and model-averaged predictive distributions. This article proposes a two-level generalization of SSVS to account for missing predictors while accommodating uncertainty in the relationships between these predictors. Bayesian approaches for allowing predictors that are missing at random require a model on the joint distribution of the predictors. We show that predictive performance can be improved by allowing uncertainty in the specification of predictor relationships in this model. The methods are illustrated through simulation studies and analysis of an epidemiologic data set.
This record has no associated files available for download.
More information
Published date: January 2010
Organisations:
Statistics
Identifiers
Local EPrints ID: 181607
URI: http://eprints.soton.ac.uk/id/eprint/181607
ISSN: 1557-4679
PURE UUID: 0ae93c11-d821-4232-b629-0cf0e1918c13
Catalogue record
Date deposited: 19 Apr 2011 10:34
Last modified: 14 Mar 2024 02:56
Export record
Altmetrics
Contributors
Author:
David Dunson
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics