Quantifying uncertainty in transdimensional Markov chain Monte Carlo using discrete Markov models

DOI10.1007/S11222-018-9828-0zbMATH Open1430.62056arXiv1703.10364OpenAlexW2605246668WikidataQ129384244 ScholiaQ129384244MaRDI QIDQ142140FDOQ142140

Quentin F. Gronau, Eric-Jan Wagenmakers, Antony M. Overstall, Eric-Jan Wagenmakers, Daniel W. Heck, Antony M. Overstall, Quentin F. Gronau

Publication date: 9 August 2018

Published in: Statistics and Computing (Search for Journal in Brave)

Abstract: Bayesian analysis often concerns an evaluation of models with different dimensionality as is necessary in, for example, model selection or mixture models. To facilitate this evaluation, transdimensional Markov chain Monte Carlo (MCMC) relies on sampling a discrete indexing variable to estimate the posterior model probabilities. However, little attention has been paid to the precision of these estimates. If only few switches occur between the models in the transdimensional MCMC output, precision may be low and assessment based on the assumption of independent samples misleading. Here, we propose a new method to estimate the precision based on the observed transition matrix of the model-indexing variable. Assuming a first order Markov model, the method samples from the posterior of the stationary distribution. This allows assessment of the uncertainty in the estimated posterior model probabilities, model ranks, and Bayes factors. Moreover, the method provides an estimate for the effective sample size of the MCMC output. In two model-selection examples, we show that the proposed approach provides a good assessment of the uncertainty associated with the estimated posterior model probabilities.

Full work available at URL: https://arxiv.org/abs/1703.10364

Recommendations

• Trans-dimensional Markov chains and their applications in statistics
• A geometric approach to transdimensional markov chain monte carlo
• Transdimensional transformation based Markov chain Monte Carlo
• Reversible jump Markov chain Monte Carlo computation and Bayesian model determination
• Population-Based Reversible Jump Markov Chain Monte Carlo

zbMATH Keywords

Bayes factorBayesian model selectionposterior model probabilitiesproduct space MCMCreversible jump MCMC

Mathematics Subject Classification ID

Bayesian inference (62F15) Numerical analysis or methods applied to Markov chains (65C40) Markov chains (discrete-time Markov processes on discrete state spaces) (60J10)

Cites Work

• Reversible jump Markov chain Monte Carlo computation and Bayesian model determination
• RcppArmadillo: accelerating R with high-performance C++ linear algebra
• Markov chain Monte Carlo Estimation of Classical and Dynamic Switching and Mixture Models
• A Markovian Function of a Markov Chain
• A fast procedure for model search in multidimensional contingency tables
• Bayes Factors
• Title not available (Why is that?)
• Markov chain Monte Carlo model determination for hierarchical and graphical log-linear models
• Title not available (Why is that?)
• Transdimensional Markov Chains
• Bayes and empirical-Bayes multiplicity adjustment in the variable-selection problem
• A tutorial on Bayes factor estimation with the product space method
• Title not available (Why is that?)
• Bayesian analysis of mixture models with an unknown number of components\,--\,an alternative to reversible jump methods.
• Bayesian variable and link determination for generalised linear models
• Statistical Inference about Markov Chains
• On Bayesian model and variable selection using MCMC
• Markov chain Monte Carlo estimation of quantiles
• A default prior distribution for contingency tables with dependent factor levels
• Variable selection for regression models
• Fixed-Width Output Analysis for Markov Chain Monte Carlo
• Relative fixed-width stopping rules for Markov chain Monte Carlo simulations
• Capture-Recapture Estimation Using Finite Mixtures of Arbitrary Dimension
• Title not available (Why is that?)
• Efficient Construction of Reversible Jump Markov Chain Monte Carlo Proposal Distributions
• Bayesian Multimodel Inference by RJMCMC: A Gibbs Sampling Approach
• What Does Objective Mean in a Dirichlet‐multinomial Process?
• Reversible jump methods for generalised linear models and generalised linear mixed models

Cited In (4)

• Multinomial models with linear inequality constraints: overview and improvements of computational methods for Bayesian inference
• Analyzing Markov chain Monte Carlo output
• A convergence diagnostic for Bayesian clustering
• MCMCprecision

Uses Software

• GLIM
• CODA
• Armadillo
• JAGS
• RcppArmadillo
• MCMCprecision
• rjmcmc
• conting
• gvs_BUGS

This page was built for publication: Quantifying uncertainty in transdimensional Markov chain Monte Carlo using discrete Markov models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q142140)