|
Activity Number:
|
37
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 2, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #304237 |
|
Title:
|
Fast Bayesian Analysis of Spatial Dynamic Factor Models
|
|
Author(s):
|
Christopher M. Strickland*+ and Daniel Simpson and Ian Turner and Robert Denham and Kerrie Mengersen
|
|
Companies:
|
Queensland University of Technology and Queensland University of Technology and Queensland University of Technology and Department of Natural Resources and Queensland University of Technology
|
|
Address:
|
2 George Street, Brisbane, International, 4001, Australia
|
|
Keywords:
|
Bayesian analysis ; Spatial dynamic factor model ; Gaussian Markov random field ; Krylov subspace method ; MODIS ; Markov chain Monte Carlo
|
|
Abstract:
|
A Bayesian Markov chain Monte Carlo (MCMC) algorithm is proposed for the efficient estimation of spatial dynamic factor models (DFMs). The spatial DFM is specified whereby spatial dependence is modeled though the columns of the factor loadings matrix using a Gaussian Markov random field. Krylov subspace methods are used to take advantage of the sparse matrix structures that are inherent in the model. The methodology is used to analyze remotely sensed data from the Moderate Imaging Spectroradiometer satellite. The data set focuses on a region in central Queensland, Australia, which contains two land type classes. The spatial DFM is used to extract both the land type information and the associated common factors in the analysis.
|
