Variance analysis of multiple importance sampling schemes

DOI10.1080/02331888.2023.2294484OpenAlexMaRDI QIDQ6192201FDO

Authors Rahul Mukerjee, Víctor Elvira

Publication date 12 February 2024

Published in Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2207.04187

zbMATH Keywords

Monte Carlo methodsselection strategyvariance inequalitiesproposal density

Mathematics Subject Classification ID

Monte Carlo methods (65C05) Statistical sampling theory and related topics (62D99)

Abstract: Multiple importance sampling (MIS) is an increasingly used methodology where several proposal densities are used to approximate integrals, generally involving target probability density functions. The use of several proposals allows for a large variety of sampling and weighting schemes. Then, the practitioner must choose a given scheme, i.e., sampling mechanism and weighting function. A variance analysis has been proposed in Elvira et al (2019, Statistical Science 34, 129-155), showing the superiority of the balanced heuristic estimator with respect to other competing schemes in some scenarios. However, some of their results are valid only for two proposals. In this paper, we extend and generalize these results, providing novel proofs that allow to determine the variance relations among MIS schemes.

Cites work

• scientific article; zbMATH DE number 2117879 (Why is no real title available?)
• An Adaptive Population Importance Sampler: Learning From Uncertainty
• Convergence rates for optimised adaptive importance samplers
• Generalized multiple importance sampling
• Importance sampling: intrinsic dimension and computational cost
• Langevin incremental mixture importance sampling
• Layered adaptive importance sampling
• Probability Inequalities for Sums of Bounded Random Variables
• Safe and Effective Importance Sampling
• The sample size required in importance sampling

This page was built for publication: Variance analysis of multiple importance sampling schemes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6192201)