Information geometry for approximate Bayesian computation

DOI10.1137/18M123284XMaRDI QIDQ4960991zbMATH OpenOpenAlexFDO

Authors Konstantinos Spiliopoulos

Publication date 24 April 2020

Published in SIAM/ASA Journal on Uncertainty Quantification (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1812.02127

zbMATH Keywords

ABCBayesianinformation theoryrelative entropyuncertainty quantification

Mathematics Subject Classification ID

Applications of renewal theory (reliability, demand theory, etc.) (60K10) Information geometry (statistical aspects) (62B11) Asymptotic properties of parametric estimators (62F12) Bayesian inference (62F15) Monte Carlo methods (65C05)

Abstract: The goal of this paper is to explore the basic Approximate Bayesian Computation (ABC) algorithm via the lens of information theory. ABC is a widely used algorithm in cases where the likelihood of the data is hard to work with or intractable, but one can simulate from it. We use relative entropy ideas to analyze the behavior of the algorithm as a function of the threshold parameter and of the size of the data. Relative entropy here is data driven as it depends on the values of the observed statistics. Relative entropy also allows us to explore the effect of the distance metric and sets up a mathematical framework for sensitivity analysis allowing to find important directions which could lead to lower computational cost of the algorithm for the same level of accuracy. In addition, we also investigate the bias of the estimators for generic observables as a function of both the threshold parameters and the size of the data. Our analysis provides error bounds on performance for positive tolerances and finite sample sizes. Simulation studies complement and illustrate the theoretical results.

Recommendations

• The rate of convergence for approximate Bayesian computation
• New insights into approximate Bayesian computation
• An Introduction to Approximate Bayesian Computation
• Asymptotic properties of approximate Bayesian computation
• Adaptive approximate Bayesian computation for complex models

Cites work

• A tutorial on approximate Bayesian computation
• Approximate Bayesian computational methods
• HIV with contact tracing: a case study in approximate Bayesian computation
• scientific article; zbMATH DE number 2107836 (Why is no real title available?)
• scientific article; zbMATH DE number 3281211 (Why is no real title available?)
• Inference for Stereological Extremes
• Integrating approximate Bayesian computation with complex agent-based models for cancer research
• New insights into approximate Bayesian computation
• Nonlinear programming
• Parallelization, processor communication and error analysis in lattice kinetic Monte Carlo
• Path-space information bounds for uncertainty quantification and sensitivity analysis of stochastic dynamics
• Scalable information inequalities for uncertainty quantification
• The rate of convergence for approximate Bayesian computation

Cited in

(4)

• Information geometry of estimating functions in semi-parametric statistical models
• Distance-learning for approximate Bayesian computation to model a volcanic eruption
• Learning functions and approximate Bayesian computation design: ABCD
• Bayesian or Laplacien inference, entropy and information theory and information geometry in data and signal processing

Describes a project that uses

Uses Software

• GPS-ABC

This page was built for publication: Information geometry for approximate Bayesian computation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4960991)