Considerate approaches to constructing summary statistics for ABC model selection
From MaRDI portal
Publication:693357
Abstract: For nearly any challenging scientific problem evaluation of the likelihood is problematic if not impossible. Approximate Bayesian computation (ABC) allows us to employ the whole Bayesian formalism to problems where we can use simulations from a model, but cannot evaluate the likelihood directly. When summary statistics of real and simulated data are compared --- rather than the data directly --- information is lost, unless the summary statistics are sufficient. Here we employ an information-theoretical framework that can be used to construct (approximately) sufficient statistics by combining different statistics until the loss of information is minimized. Such sufficient sets of statistics are constructed for both parameter estimation and model selection problems. We apply our approach to a range of illustrative and real-world model selection problems.
Recommendations
- Constructing summary statistics for approximate Bayesian computation: semi-automatic approximate Bayesian computation. With discussion and authors' reply
- Semi-automatic selection of summary statistics for ABC model choice
- On optimal selection of summary statistics for approximate Bayesian computation
- Choosing summary statistics by least angle regression for approximate Bayesian computation
- Deviance information criteria for model selection in approximate Bayesian computation
Cites work
- scientific article; zbMATH DE number 2126323 (Why is no real title available?)
- scientific article; zbMATH DE number 3522963 (Why is no real title available?)
- scientific article; zbMATH DE number 1220667 (Why is no real title available?)
- scientific article; zbMATH DE number 2061729 (Why is no real title available?)
- Approximately sufficient statistics and Bayesian computation
- Bayesian data analysis.
- Elements of Information Theory
- Elements of the Random Walk
- Information, Physics, and Computation
- Likelihood-free estimation of model evidence
- Mathematical Statistics
- Mathematical population genetics. I: Theoretical introduction.
- Model Selection and Multimodel Inference
- On approximate sufficiency
- On optimal selection of summary statistics for approximate Bayesian computation
- Parameter estimation for hidden Markov models with intractable likelihoods
- Principles of Statistical Inference
- The Bayesian Choice
Cited in
(14)- Goodness of fit for models with intractable likelihood
- Deviance information criteria for model selection in approximate Bayesian computation
- On selection of statistics for approximate Bayesian computing (or the method of simulated moments)
- Semi-automatic selection of summary statistics for ABC model choice
- Approximate Bayesian computation by modelling summary statistics in a quasi-likelihood framework
- Approximate Bayesian computation and simulation-based inference for complex stochastic epidemic models
- On optimal selection of summary statistics for approximate Bayesian computation
- Ancestral inference in tumors: how much can we know?
- A comparative review of dimension reduction methods in approximate Bayesian computation
- Spectral density-based and measure-preserving ABC for partially observed diffusion processes. An illustration on Hamiltonian SDEs
- Weak convergence of posteriors conditional on maximum pseudo-likelihood estimates and implications in ABC
- Open problems in mathematical biology
- Rapid Bayesian Inference for Expensive Stochastic Models
- Adaptive ABC model choice and geometric summary statistics for hidden Gibbs random fields
This page was built for publication: Considerate approaches to constructing summary statistics for ABC model selection
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q693357)
