Adaptive ABC model choice and geometric summary statistics for hidden Gibbs random fields

DOI10.1007/S11222-014-9514-9zbMATH Open1331.62066arXiv1402.1380OpenAlexW3100993489MaRDI QIDQ5963545FDOQ5963545

Pierre Pudlo, Lionel Cucala, Julien Stoehr

Publication date: 22 February 2016

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

Abstract: Selecting between different dependency structures of hidden Markov random field can be very challenging, due to the intractable normalizing constant in the likelihood. We answer this question with approximate Bayesian computation (ABC) which provides a model choice method in the Bayesian paradigm. This comes after the work of Grelaud et al. (2009) who exhibited sufficient statistics on directly observed Gibbs random fields. But when the random field is latent, the sufficiency falls and we complement the set with geometric summary statistics. The general approach to construct these intuitive statistics relies on a clustering analysis of the sites based on the observed colors and plausible latent graphs. The efficiency of ABC model choice based on these statistics is evaluated via a local error rate which may be of independent interest. As a byproduct we derived an ABC algorithm that adapts the dimension of the summary statistics to the dataset without distorting the model selection.

Full work available at URL: https://arxiv.org/abs/1402.1380

zbMATH Keywords

approximate Bayesian computationmodel choicemisclassification ratesummary statistics\(k\)-nearest neighborshidden Gibbs random fields

Mathematics Subject Classification ID

Nonparametric estimation (62G05) Markov processes: estimation; hidden Markov models (62M05) Random fields; image analysis (62M40) Bayesian problems; characterization of Bayes procedures (62C10)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• A comparative review of dimension reduction methods in approximate Bayesian computation
• Perfect slice samplers
• Pre-processing for approximate Bayesian computation in image analysis
• Approximate Bayesian computational methods
• Adaptive approximate Bayesian computation
• ABC likelihood-free methods for model choice in Gibbs random fields
• Hidden Markov Models and Disease Mapping
• An adaptive sequential Monte Carlo method for approximate Bayesian computation
• Invariant HPD credible sets and MAP estimators
• Likelihood-free estimation of model evidence
• Recursive computing and simulation-free inference for general factorizable models
• Efficient recursions for general factorisable models
• An overview on Approximate Bayesian computation

Cited In (5)

• Prior specification for binary Markov mesh models
• Conditional random fields for pattern recognition applied to structured data
• Approximate Bayesian Computation: A Survey on Recent Results
• Pre-processing for approximate Bayesian computation in image analysis
• Hidden Gibbs random fields model selection using block likelihood information criterion

This page was built for publication: Adaptive ABC model choice and geometric summary statistics for hidden Gibbs random fields

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5963545)