Automated parameter blocking for efficient Markov chain Monte Carlo sampling

DOI10.1214/16-BA1008MaRDI QIDQ1699668zbMATH OpenOpenAlexFDO

Authors Daniel Turek, Perry de Valpine, Christopher J. Paciorek, Clifford Anderson-Bergman

Publication date 23 February 2018

Published in Bayesian Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1503.05621, https://projecteuclid.org/euclid.ba/1464266500

zbMATH Keywords

block samplingintegrated autocorrelation timeMCMCMetropolis-HastingsmixingNIMBLE

Mathematics Subject Classification ID

General considerations in statistical decision theory (62C05) Numerical analysis or methods applied to Markov chains (65C40)

Abstract: Markov chain Monte Carlo (MCMC) sampling is an important and commonly used tool for the analysis of hierarchical models. Nevertheless, practitioners generally have two options for MCMC: utilize existing software that generates a black-box "one size fits all" algorithm, or the challenging (and time consuming) task of implementing a problem-specific MCMC algorithm. Either choice may result in inefficient sampling, and hence researchers have become accustomed to MCMC runtimes on the order of days (or longer) for large models. We propose an automated procedure to determine an efficient MCMC algorithm for a given model and computing platform. Our procedure dynamically determines blocks of parameters for joint sampling that result in efficient sampling of the entire model. We test this procedure using a diverse suite of example models, and observe non-trivial improvements in MCMC efficiency for many models. Our procedure is the first attempt at such, and may be generalized to a broader space of MCMC algorithms. Our results suggest that substantive improvements in MCMC efficiency may be practically realized using our automated blocking procedure, or variants thereof, which warrants additional study and application.

Recommendations

• Efficient MCMC sampling in dynamic mixture models
• An automated (Markov chain) Monte Carlo EM algorithm
• Nested adaptation of MCMC algorithms
• The no-U-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo
• scientific article; zbMATH DE number 849925

Cited in

(10)

• Nested adaptation of MCMC algorithms
• Spatiotemporal blocking of the bouncy particle sampler for efficient inference in state-space models
• Block Gibbs samplers for logistic mixed models: convergence properties and a comparison with full Gibbs samplers
• On the convergence rate of the ``out-of-order block Gibbs sampler
• Demand Models With Random Partitions
• Simulating Markov Random Fields With a Conclique-Based Gibbs Sampler
• Bayesian Pairwise Comparison of High-Dimensional Images
• Unconstrained Metropolis-Hastings sampling of covariance matrices
• A hybrid Markov chain Monte Carlo approach for structural learning in Bayesian networks based on variable blocking
• Efficient MCMC sampling in dynamic mixture models

Describes a project that uses

Uses Software

• WinBUGS
• nimble
• R
• JAGS
• BUGS
• Eigen
• Stan

This page was built for publication: Automated parameter blocking for efficient Markov chain Monte Carlo sampling

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1699668)