Generalized Multiple Importance Sampling

Abstract

Importance Sampling methods are broadly used to approximate posterior distributions or some of their moments. In its standard approach, samples are drawn from a single proposal distribution and weighted properly. However, since the performance depends on the mismatch between the targeted and the proposal distributions, several proposal densities are often employed for the generation of samples. Under this Multiple Importance Sampling (MIS) scenario, many works have addressed the selection or adaptation of the proposal distributions, interpreting the sampling and the weighting steps in different ways. In this paper, we establish a general framework for sampling and weighing procedures when more than one proposal are available. The most relevant MIS schemes in the literature are encompassed within the new framework, and, moreover novel valid schemes appear naturally. All the MIS schemes are compared and ranked in terms of the variance of the associated estimators. Finally, we provide illustrative examples which reveal that, even with a good choice of the proposal densities, a careful interpretation of the sampling and weighting procedures can make a significant difference in the performance of the method.

References in corpus (7)

• Adaptive Importance Sampling in General Mixture Classes
• Layered Adaptive Importance Sampling
• Improving Population Monte Carlo: Alternative Weighting and Resampling Schemes
• Convergence of adaptive mixtures of importance sampling schemes
• Efficient Multiple Importance Sampling Estimators
• Heretical Multiple Importance Sampling
• Optimal mixture weights in multiple importance sampling

Cited by in corpus (36)

• Effective Sample Size for Importance Sampling based on discrepancy measures
• Layered Adaptive Importance Sampling
• Improving Population Monte Carlo: Alternative Weighting and Resampling Schemes
• Trajectory Poisson multi-Bernoulli filters
• Marginal likelihood computation for model selection and hypothesis testing: an extensive review
• Rethinking the Effective Sample Size
• Group Importance Sampling for Particle Filtering and MCMC
• Importance Gaussian Quadrature
• Implicitly Adaptive Importance Sampling
• Heretical Multiple Importance Sampling
• Robust Covariance Adaptation in Adaptive Importance Sampling
• Optimized Population Monte Carlo
• Issues in the Multiple Try Metropolis mixing
• Pathfinder: Parallel quasi-Newton variational inference
• Hamiltonian Adaptive Importance Sampling
• MCMC-driven importance samplers
• Overdispersed Black-Box Variational Inference
• Asymptotic optimality of adaptive importance sampling
• Rare Events via Cross-Entropy Population Monte Carlo
• Warp Bridge Sampling: The Next Generation
• Safe and adaptive importance sampling: a mixture approach
• Consistent Adaptive Multiple Importance Sampling and Controlled Diffusions
• Filtering Variational Objectives
• Tree pyramidal adaptive importance sampling
• Population Quasi-Monte Carlo
• Rao-Blackwellization in the MCMC era
• Optimized Auxiliary Particle Filters: adapting mixture proposals via convex optimization
• Deep Importance Sampling based on Regression for Model Inversion and Emulation
• Path Throughput Importance Weights
• Adaptive posterior distributions for uncertainty analysis of covariance matrices in Bayesian inversion problems for multioutput signals
• A Sequential Importance Sampling Algorithm for Estimating Linear Extensions
• Revisiting the balance heuristic for estimating normalising constants
• Advances in Importance Sampling
• Estimation and prediction for spatial generalized linear mixed models with parametric links via reparameterized importance sampling
• Auto-encoder model for faster generation of effective one-body gravitational waveform approximations
• Bayesian Fusion of Data Partitioned Particle Estimates