Two-stage Metropolis-Hastings for tall data

DOI10.1007/S00357-018-9248-ZzbMATH Open1391.62120DBLPjournals/classification/PayneM18arXiv1411.5653OpenAlexW2625684004WikidataQ57146845 ScholiaQ57146845MaRDI QIDQ724596FDOQ724596

Authors: Richard D. Payne, Bani Mallick 

Publication date: 26 July 2018

Published in: Journal of Classification (Search for Journal in Brave)

Abstract: This paper discusses the challenges presented by tall data problems associated with Bayesian classification (specifically binary classification) and the existing methods to handle them. Current methods include parallelizing the likelihood, subsampling, and consensus Monte Carlo. A new method based on the two-stage Metropolis-Hastings algorithm is also proposed. The purpose of this algorithm is to reduce the exact likelihood computational cost in the tall data situation. In the first stage, a new proposal is tested by the approximate likelihood based model. The full likelihood based posterior computation will be conducted only if the proposal passes the first stage screening. Furthermore, this method can be adopted into the consensus Monte Carlo framework. The two-stage method is applied to logistic regression, hierarchical logistic regression, and Bayesian multivariate adaptive regression splines.

Full work available at URL: https://arxiv.org/abs/1411.5653

Recommendations

• scientific article; zbMATH DE number 6781368
• Extended stochastic gradient Markov chain Monte Carlo for large-scale Bayesian variable selection
• Double-parallel Monte Carlo for Bayesian analysis of big data
• Parallel Markov chain Monte Carlo for Bayesian hierarchical models with big data, in two stages
• Comparing consensus Monte Carlo strategies for distributed Bayesian computation

zbMATH Keywords

Markov chain Monte CarloBayesian inferenceMetropolis-Hastings algorithmlogistic modelBayesian multivariate adaptive regression splinestall data

Mathematics Subject Classification ID

Bayesian inference (62F15) Classification and discrimination; cluster analysis (statistical aspects) (62H30)

Cites Work

• Reversible jump Markov chain Monte Carlo computation and Bayesian model determination
• Multivariate adaptive regression splines
• Title not available (Why is that?)
• The pseudo-marginal approach for efficient Monte Carlo computations
• 10.1162/15324430152748236
• Speeding Up MCMC by Efficient Data Subsampling
• Title not available (Why is that?)
• Bayesian Classification of Tumours by Using Gene Expression Data
• A Bayesian approach to characterizing uncertainty in inverse problems using coarse and fine-scale information
• Generalized Nonlinear Modeling With Multivariate Free-Knot Regression Splines
• Classification with Bayesian MARS

Cited In (6)

• Speeding up MCMC by Delayed Acceptance and Data Subsampling
• The Block-Poisson Estimator for Optimally Tuned Exact Subsampling MCMC
• Scalable Bayesian Nonparametric Clustering and Classification
• Accelerating sequential Monte Carlo with surrogate likelihoods
• A two-stage adaptive Metropolis algorithm
• Title not available (Why is that?)

This page was built for publication: Two-stage Metropolis-Hastings for tall data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q724596)