Journal article
Independent sampling for Bayesian normal conditional autoregressive models with OpenCL acceleration
Computational Statistics, Vol.33(1), pp.159-177
03/2018
DOI: 10.1007/s00180-017-0752-0
Abstract
A new computational strategy produces independent samples from the joint posterior distribution for a broad class of Bayesian spatial and spatiotemporal conditional autoregressive models. The method is based on reparameterization and marginalization of the posterior distribution and massive parallelization of rejection sampling using graphical processing units (GPUs) or other accelerators. It enables very fast sampling for small to moderate-sized datasets (up to approximately 10,000 observations) and feasible sampling for much larger datasets. Even using a mid-range GPU and a high-end CPU, the GPU-based implementation is up to 30 times faster than the same algorithm run serially on a single CPU, and the numbers of effective samples per second are orders of magnitude higher than those obtained with popular Markov chain Monte Carlo software. The method has been implemented in the R package CARrampsOcl. This work provides both a practical computing strategy for fitting a popular class of Bayesian models and a proof of concept that GPU acceleration can make independent sampling from Bayesian joint posterior densities feasible.
Details
- Title: Subtitle
- Independent sampling for Bayesian normal conditional autoregressive models with OpenCL acceleration
- Creators
- Mary Cowles - Department of Statistics and Actuarial Science The University of Iowa Iowa City IA 52242 USAStephen Bonett - University of Pennsylvania School of Nursing 418 Curie Blvd. Philadelphia PA 19104-4217 USAMichael Seedorff - Department of Biostatistics The University of Iowa Iowa City IA 52242 USA
- Resource Type
- Journal article
- Publication Details
- Computational Statistics, Vol.33(1), pp.159-177
- Publisher
- Springer Berlin Heidelberg; Berlin/Heidelberg
- DOI
- 10.1007/s00180-017-0752-0
- ISSN
- 0943-4062
- eISSN
- 1613-9658
- Language
- English
- Date published
- 03/2018
- Academic Unit
- Biostatistics; Statistics and Actuarial Science
- Record Identifier
- 9983985701702771
Metrics
23 Record Views