Variational approximation for importance sampling

Xiao Su, Yuguo Chen

Research output: Contribution to journalArticlepeer-review

Abstract

We propose an importance sampling algorithm with proposal distribution obtained from variational approximation. This method combines the strength of both importance sampling and variational method. On one hand, this method avoids the bias from variational method. On the other hand, variational approximation provides a way to design the proposal distribution for the importance sampling algorithm. Theoretical justification of the proposed method is provided. Numerical results show that using variational approximation as the proposal can improve the performance of importance sampling and sequential importance sampling.

Original languageEnglish (US)
Pages (from-to)1901-1930
Number of pages30
JournalComputational Statistics
Volume36
Issue number3
DOIs
StateAccepted/In press - 2021

Keywords

  • Monte Carlo
  • Proposal distribution
  • Variational inference
  • f-divergence

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty
  • Computational Mathematics

Fingerprint

Dive into the research topics of 'Variational approximation for importance sampling'. Together they form a unique fingerprint.

Cite this