Bayesian Analysis
- Bayesian Anal.
- Volume 13, Number 2 (2018), 437-459.
Efficient Model Comparison Techniques for Models Requiring Large Scale Data Augmentation
Panayiota Touloupou, Naif Alzahrani, Peter Neal, Simon E. F. Spencer, and Trevelyan J. McKinley
Abstract
Selecting between competing statistical models is a challenging problem especially when the competing models are non-nested. In this paper we offer a simple solution by devising an algorithm which combines MCMC and importance sampling to obtain computationally efficient estimates of the marginal likelihood which can then be used to compare the models. The algorithm is successfully applied to a longitudinal epidemic data set, where calculating the marginal likelihood is made more challenging by the presence of large amounts of missing data. In this context, our importance sampling approach is shown to outperform existing methods for computing the marginal likelihood.
Article information
Source
Bayesian Anal., Volume 13, Number 2 (2018), 437-459.
Dates
First available in Project Euclid: 29 April 2017
Permanent link to this document
https://projecteuclid.org/euclid.ba/1493431262
Digital Object Identifier
doi:10.1214/17-BA1057
Mathematical Reviews number (MathSciNet)
MR3780430
Zentralblatt MATH identifier
06989955
Keywords
epidemics marginal likelihood model evidence model selection
Rights
Creative Commons Attribution 4.0 International License.
Citation
Touloupou, Panayiota; Alzahrani, Naif; Neal, Peter; Spencer, Simon E. F.; McKinley, Trevelyan J. Efficient Model Comparison Techniques for Models Requiring Large Scale Data Augmentation. Bayesian Anal. 13 (2018), no. 2, 437--459. doi:10.1214/17-BA1057. https://projecteuclid.org/euclid.ba/1493431262
Supplemental materials
- Supplementary material: Efficient model comparison techniques for models requiring large scale data augmentation. Digital Object Identifier: doi:10.1214/17-BA1057SUPP