Diagnostic tools for approximate Bayesian computation using the coverage property
From MaRDI portal
Publication:2804160
DOI10.1111/ANZS.12087zbMATH Open1335.62058arXiv1301.3166OpenAlexW2594309842MaRDI QIDQ2804160FDOQ2804160
G. Popovic, S. A. Sisson, D. Prangle, M. G. B. Blum
Publication date: 28 April 2016
Published in: Australian \& New Zealand Journal of Statistics (Search for Journal in Brave)
Abstract: Approximate Bayesian computation (ABC) is an approach for sampling from an approximate posterior distribution in the presence of a computationally intractable likelihood function. A common implementation is based on simulating model, parameter and dataset triples, (m, heta,y), from the prior, and then accepting as samples from the approximate posterior, those pairs (m, heta) for which y, or a summary of y, is "close" to the observed data. Closeness is typically determined though a distance measure and a kernel scale parameter, epsilon. Appropriate choice of epsilon is important to producing a good quality approximation. This paper proposes diagnostic tools for the choice of epsilon based on assessing the coverage property, which asserts that credible intervals have the correct coverage levels. We provide theoretical results on coverage for both model and parameter inference, and adapt these into diagnostics for the ABC context. We re-analyse a study on human demographic history to determine whether the adopted posterior approximation was appropriate. R code implementing the proposed methodology is freely available in the package "abc."
Full work available at URL: https://arxiv.org/abs/1301.3166
Bayesian inference (62F15) Diagnostics, and linear inference and regression (62J20) Software, source code, etc. for problems pertaining to statistics (62-04) Applications of statistics to social sciences (62P25)
Cites Work
- Title not available (Why is that?)
- Probabilistic Forecasts, Calibration and Sharpness
- A comparative review of dimension reduction methods in approximate Bayesian computation
- Present Position and Potential Developments: Some Personal Views: Statistical Theory: The Prequential Approach
- Approximate Bayesian Computation: A Nonparametric Perspective
- Inference for Stereological Extremes
- Approximate Bayesian computational methods
- ABC likelihood-free methods for model choice in Gibbs random fields
- Likelihood-free Bayesian estimation of multivariate quantile distributions
- Parameter Estimation for Hidden Markov Models with Intractable Likelihoods
- Proper likelihoods for Bayesian analysis
- On Testing the Validity of Sequential Probability Forecasts
- Simultaneous adjustment of bias and coverage probabilities for confidence intervals
Cited In (17)
- Approximate Bayesian inference for agent-based models in economics: a case study
- Extending approximate Bayesian computation methods to high dimensions via a Gaussian copula model
- Scalable Bayesian inference for the inverse temperature of a hidden Potts model
- Calibration procedures for approximate Bayesian credible sets
- Simultaneous adjustment of bias and coverage probabilities for confidence intervals
- An approximate likelihood perspective on ABC methods
- Recalibration: a post-processing method for approximate Bayesian computation
- Debiasing inference with approximate covariance matrices and other unidentified biases
- SB-ETAS: using simulation based inference for scalable, likelihood-free inference for the ETAS model of earthquake occurrences
- Approximate Bayesian Computation: A Survey on Recent Results
- A quasi likelihood approximation of posterior distributions for likelihood-intractable complex models
- ABC–CDE: Toward Approximate Bayesian Computation With Complex High-Dimensional Data and Limited Simulations
- An improved approximate Bayesian computation scheme for parameter inference based on a recalibration post-processing method
- Assessment and Adjustment of Approximate Inference Algorithms Using the Law of Total Variance
- Spectral density-based and measure-preserving ABC for partially observed diffusion processes. An illustration on Hamiltonian SDEs
- Likelihood-free approximate Gibbs sampling
- Approximate Bayesian computation by modelling summary statistics in a quasi-likelihood framework
Uses Software
This page was built for publication: Diagnostic tools for approximate Bayesian computation using the coverage property
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2804160)
