Pre-processing for approximate Bayesian computation in image analysis

DOI10.1007/S11222-014-9525-6zbMATH Open1331.62158arXiv1403.4359OpenAlexW2118661871WikidataQ56994086 ScholiaQ56994086MaRDI QIDQ5963548FDOQ5963548

Matthew T. Moores, C. C. Drovandi, Christian Robert, K. Mengersen

Publication date: 22 February 2016

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

Abstract: Most of the existing algorithms for approximate Bayesian computation (ABC) assume that it is feasible to simulate pseudo-data from the model at each iteration. However, the computational cost of these simulations can be prohibitive for high dimensional data. An important example is the Potts model, which is commonly used in image analysis. Images encountered in real world applications can have millions of pixels, therefore scalability is a major concern. We apply ABC with a synthetic likelihood to the hidden Potts model with additive Gaussian noise. Using a pre-processing step, we fit a binding function to model the relationship between the model parameters and the synthetic likelihood parameters. Our numerical experiments demonstrate that the precomputed binding function dramatically improves the scalability of ABC, reducing the average runtime required for model fitting from 71 hours to only 7 minutes. We also illustrate the method by estimating the smoothing parameter for remotely sensed satellite imagery. Without precomputation, Bayesian inference is impractical for datasets of that scale.

Full work available at URL: https://arxiv.org/abs/1403.4359

zbMATH Keywords

sequential Monte Carloapproximate Bayesian computationindirect inferencequasi-likelihoodhidden Markov random fieldPotts/Ising model

Mathematics Subject Classification ID

Bayesian inference (62F15) Bayesian problems; characterization of Bayes procedures (62C10) Image analysis in multivariate analysis (62H35) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• RcppArmadillo: accelerating R with high-performance C++ linear algebra
• Bayesian indirect inference using a parametric auxiliary model
• Auxiliary Variable Methods for Markov Chain Monte Carlo with Applications
• Sequential Monte Carlo without likelihoods
• Monte Carlo strategies in scientific computing
• Adaptive approximate Bayesian computation
• ABC likelihood-free methods for model choice in Gibbs random fields
• Estimation of Parameters for Macroparasite Population Evolution Using Approximate Bayesian Computation
• An adaptive sequential Monte Carlo method for approximate Bayesian computation
• Filtering via approximate Bayesian computation
• A quasi likelihood approximation of posterior distributions for likelihood-intractable complex models
• Encompassing and indirect inference
• A Bayesian Reassessment of Nearest-Neighbor Classification
• Classification using distance nearest neighbours
• Adaptive ABC model choice and geometric summary statistics for hidden Gibbs random fields

Cited In (19)

• Scalable Bayesian inference for the inverse temperature of a hidden Potts model
• Variational Bayes with synthetic likelihood
• Likelihood-free inference in high dimensions with synthetic likelihood
• An approximate likelihood perspective on ABC methods
• Approximation of Bayesian predictive \(p\)-values with regression ABC
• A synthetic likelihood approach for intractable Markov random fields
• Bayesian computational methods for spatial analysis of images. (Abstract of thesis)
• A Function Emulation Approach for Doubly Intractable Distributions
• Lazy ABC
• Bayesian Synthetic Likelihood
• Accuracy of the Bethe approximation for hyperparameter estimation in probabilistic image processing
• Title not available (Why is that?)
• Spectral density-based and measure-preserving ABC for partially observed diffusion processes. An illustration on Hamiltonian SDEs
• Efficient MCMC for Gibbs random fields using pre-computation
• Scalable approximate Bayesian computation for growing network models via extrapolated and sampled summaries
• On Russian roulette estimates for Bayesian inference with doubly-intractable likelihoods
• Adaptive ABC model choice and geometric summary statistics for hidden Gibbs random fields
• Introduction to ``Pre-processing for approximate Bayesian computation in image analysis by M. Moores, C. Drovandi, K. Mengersen, C. Robert
• Bayesian indirect inference for models with intractable normalizing functions

Uses Software

• RcppArmadillo

This page was built for publication: Pre-processing for approximate Bayesian computation in image analysis

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5963548)