Bayesian Model Selection in Spatial Lattice Models

• Activity Number: 298
• Type: Contributed
• Date/Time: Tuesday, August 2, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Joon Jin Song*+ and Victor De Oliveira
• Companies: University of Arkansas and The University of Texas at San Antonio
• Keywords: Bayes factors ; CAR models ; SAR models ; Jeffreys prior ; Spatial data ; Weight matrix

Abstract

This work describes a Bayesian approach for model selection in Gaussian conditional autoregressive models and Gaussian simultaneous autoregressive models that are commonly used to describe spatial lattice data. The approach is aimed at situations when all competing models have the same mean structure, and the model differences rely on some aspects of the covariance structure. As the selection criterion the method uses posterior model probabilities computed using some default priors on the model parameters. The proposed method is illustrated using two real datasets.