Back to Search
Start Over
Robust Bayesian model selection for heavy-tailed linear regression using finite mixtures
- Source :
- Braz. J. Probab. Stat. 34, no. 1 (2020), 51-70
- Publication Year :
- 2020
- Publisher :
- Institute of Mathematical Statistics, 2020.
-
Abstract
- In this paper we present a novel methodology to perform Bayesian model selection in linear models with heavy-tailed distributions. We consider a finite mixture of distributions to model a latent variable where each component of the mixture corresponds to one possible model within the symmetrical class of normal independent distributions. Naturally, the Gaussian model is one of the possibilities. This allows for a simultaneous analysis based on the posterior probability of each model. Inference is performed via Markov chain Monte Carlo - a Gibbs sampler with Metropolis-Hastings steps for a class of parameters. Simulated examples highlight the advantages of this approach compared to a segregated analysis based on arbitrarily chosen model selection criteria. Examples with real data are presented and an extension to censored linear regression is introduced and discussed.
- Subjects :
- FOS: Computer and information sciences
Statistics and Probability
MCMC
Posterior probability
Bayesian inference
01 natural sciences
Methodology (stat.ME)
010104 statistics & probability
symbols.namesake
Student-t
0502 economics and business
Linear regression
Applied mathematics
Scale mixtures of normal
0101 mathematics
Statistics - Methodology
050205 econometrics
Mathematics
slash
Markov chain
Model selection
05 social sciences
Linear model
Markov chain Monte Carlo
Statistics::Computation
symbols
penalised complexity priors
Gibbs sampling
Subjects
Details
- ISSN :
- 01030752
- Volume :
- 34
- Database :
- OpenAIRE
- Journal :
- Brazilian Journal of Probability and Statistics
- Accession number :
- edsair.doi.dedup.....d11330bc32ae568e29dbd9ed64d684f5
- Full Text :
- https://doi.org/10.1214/18-bjps417