Automated factor slice sampling

Matthew M. Tibbits, Chris Groendyke, Murali Haran, John C. Liechty

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Markov chain Monte Carlo (MCMC) algorithms offer a very general approach for sampling from arbitrary distributions. However, designing and tuning MCMC algorithms for each new distribution can be challenging and time consuming. It is particularly difficult to create an efficient sampler when there is strong dependence among the variables in a multivariate distribution. We describe a two-pronged approach for constructing efficient, automated MCMC algorithms: (1) we propose the "factor slice sampler," a generalization of the univariate slice sampler where we treat the selection of a coordinate basis (factors) as an additional tuning parameter, and (2) we develop an approach for automatically selecting tuning parameters to construct an efficient factor slice sampler. In addition to automating the factor slice sampler, our tuning approach also applies to the standard univariate slice samplers.We demonstrate the efficiency and general applicability of our automated MCMC algorithm with a number of illustrative examples. This article has online supplementary materials.

Original languageEnglish (US)
Pages (from-to)543-563
Number of pages21
JournalJournal of Computational and Graphical Statistics
Volume23
Issue number2
DOIs
StatePublished - Jan 1 2014

Fingerprint

Slice
Markov Chain Monte Carlo Algorithms
Parameter Tuning
Univariate
Tuning
Multivariate Distribution
Sampling
Markov chain Monte Carlo
Factors
Arbitrary
Demonstrate

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Discrete Mathematics and Combinatorics
  • Statistics, Probability and Uncertainty

Cite this

Tibbits, Matthew M. ; Groendyke, Chris ; Haran, Murali ; Liechty, John C. / Automated factor slice sampling. In: Journal of Computational and Graphical Statistics. 2014 ; Vol. 23, No. 2. pp. 543-563.
@article{5e177f9eb4464896a31f4a2ebe932216,
title = "Automated factor slice sampling",
abstract = "Markov chain Monte Carlo (MCMC) algorithms offer a very general approach for sampling from arbitrary distributions. However, designing and tuning MCMC algorithms for each new distribution can be challenging and time consuming. It is particularly difficult to create an efficient sampler when there is strong dependence among the variables in a multivariate distribution. We describe a two-pronged approach for constructing efficient, automated MCMC algorithms: (1) we propose the {"}factor slice sampler,{"} a generalization of the univariate slice sampler where we treat the selection of a coordinate basis (factors) as an additional tuning parameter, and (2) we develop an approach for automatically selecting tuning parameters to construct an efficient factor slice sampler. In addition to automating the factor slice sampler, our tuning approach also applies to the standard univariate slice samplers.We demonstrate the efficiency and general applicability of our automated MCMC algorithm with a number of illustrative examples. This article has online supplementary materials.",
author = "Tibbits, {Matthew M.} and Chris Groendyke and Murali Haran and Liechty, {John C.}",
year = "2014",
month = "1",
day = "1",
doi = "10.1080/10618600.2013.791193",
language = "English (US)",
volume = "23",
pages = "543--563",
journal = "Journal of Computational and Graphical Statistics",
issn = "1061-8600",
publisher = "American Statistical Association",
number = "2",

}

Automated factor slice sampling. / Tibbits, Matthew M.; Groendyke, Chris; Haran, Murali; Liechty, John C.

In: Journal of Computational and Graphical Statistics, Vol. 23, No. 2, 01.01.2014, p. 543-563.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Automated factor slice sampling

AU - Tibbits, Matthew M.

AU - Groendyke, Chris

AU - Haran, Murali

AU - Liechty, John C.

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Markov chain Monte Carlo (MCMC) algorithms offer a very general approach for sampling from arbitrary distributions. However, designing and tuning MCMC algorithms for each new distribution can be challenging and time consuming. It is particularly difficult to create an efficient sampler when there is strong dependence among the variables in a multivariate distribution. We describe a two-pronged approach for constructing efficient, automated MCMC algorithms: (1) we propose the "factor slice sampler," a generalization of the univariate slice sampler where we treat the selection of a coordinate basis (factors) as an additional tuning parameter, and (2) we develop an approach for automatically selecting tuning parameters to construct an efficient factor slice sampler. In addition to automating the factor slice sampler, our tuning approach also applies to the standard univariate slice samplers.We demonstrate the efficiency and general applicability of our automated MCMC algorithm with a number of illustrative examples. This article has online supplementary materials.

AB - Markov chain Monte Carlo (MCMC) algorithms offer a very general approach for sampling from arbitrary distributions. However, designing and tuning MCMC algorithms for each new distribution can be challenging and time consuming. It is particularly difficult to create an efficient sampler when there is strong dependence among the variables in a multivariate distribution. We describe a two-pronged approach for constructing efficient, automated MCMC algorithms: (1) we propose the "factor slice sampler," a generalization of the univariate slice sampler where we treat the selection of a coordinate basis (factors) as an additional tuning parameter, and (2) we develop an approach for automatically selecting tuning parameters to construct an efficient factor slice sampler. In addition to automating the factor slice sampler, our tuning approach also applies to the standard univariate slice samplers.We demonstrate the efficiency and general applicability of our automated MCMC algorithm with a number of illustrative examples. This article has online supplementary materials.

UR - http://www.scopus.com/inward/record.url?scp=84901807491&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84901807491&partnerID=8YFLogxK

U2 - 10.1080/10618600.2013.791193

DO - 10.1080/10618600.2013.791193

M3 - Article

VL - 23

SP - 543

EP - 563

JO - Journal of Computational and Graphical Statistics

JF - Journal of Computational and Graphical Statistics

SN - 1061-8600

IS - 2

ER -