Pareto Smoothed Importance Sampling

Abstract

Importance weighting is a general way to adjust Monte Carlo integration to account for draws from the wrong distribution, but the resulting estimate can be highly variable when the importance ratios have a heavy right tail. This routinely occurs when there are aspects of the target distribution that are not well captured by the approximating distribution, in which case more stable estimates can be obtained by modifying extreme importance ratios. We present a new method for stabilizing importance weights using a generalized Pareto distribution fit to the upper tail of the distribution of the simulated importance ratios. The method, which empirically performs better than existing methods for stabilizing importance sampling estimates, includes stabilized effective sample size estimates, Monte Carlo error estimates, and convergence diagnostics. The presented Pareto $\hat{k}$ finite sample convergence rate diagnostic is useful for any Monte Carlo estimator.

The final version with minor corrections. To be published in JMLR. 58 pages

References in corpus (4)

• Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC
• Normal approximation under local dependence
• Bayesian aggregation of average data: An application in drug development
• Projection predictive variable selection using Stan+R

Cited by in corpus (46)

• Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC
• Visualization in Bayesian workflow
• Statistical methods for linguistic research: Foundational Ideas - Part II
• Bayesian Item Response Modeling in R with brms and Stan
• Randomized Predictive P-values: A Versatile Model Diagnostic Tool with Unified Reference Distribution
• Projective Inference in High-dimensional Problems: Prediction and Feature Selection
• Rethinking the Effective Sample Size
• Monitoring the COVID-19 epidemic with nationwide telecommunication data
• Efficient leave-one-out cross-validation for Bayesian non-factorized normal and Student-t models
• Bayesian hierarchical stacking: Some models are (somewhere) useful
• Implicitly Adaptive Importance Sampling
• Heretical Multiple Importance Sampling
• Stacking for Non-mixing Bayesian Computations: The Curse and Blessing of Multimodal Posteriors
• Leave-One-Out Cross-Validation for Bayesian Model Comparison in Large Data
• Robust Covariance Adaptation in Adaptive Importance Sampling
• Black-box Variational Inference for Stochastic Differential Equations
• Robust leave-one-out cross-validation for high-dimensional Bayesian models
• Weak evidence for variable occultation depth of 55 Cnc e with TESS
• Predictive Bayesian selection of multistep Markov chains, applied to the detection of the hot hand and other statistical dependencies in free throws
• Pathfinder: Parallel quasi-Newton variational inference
• Some models are useful, but how do we know which ones? Towards a unified Bayesian model taxonomy
• Robust, Accurate Stochastic Optimization for Variational Inference
• A Bayesian Dirichlet Auto-Regressive Moving Average Model for Forecasting Lead Times
• BAMBI: An R package for Fitting Bivariate Angular Mixture Models
• Fast, Optimal, and Targeted Predictions using Parametrized Decision Analysis
• Quantifying sources of uncertainty in drug discovery predictions with probabilistic models
• LOO and WAIC as Model Selection Methods for Polytomous Items
• Confident Off-Policy Evaluation and Selection through Self-Normalized Importance Weighting
• Validated Variational Inference via Practical Posterior Error Bounds
• Hamiltonian Monte Carlo using an adjoint-differentiated Laplace approximation: Bayesian inference for latent Gaussian models and beyond
• Embarrassingly Parallel Inference for Gaussian Processes
• Embarrassingly parallel MCMC using deep invertible transformations
• Variational Refinement for Importance Sampling Using the Forward Kullback-Leibler Divergence
• BayesTime: Bayesian Functional Principal Components for Sparse Longitudinal Data
• Challenges and Opportunities in High-dimensional Variational Inference
• Projection predictive variable selection for discrete response families with finite support
• Adaptive Path Sampling in Metastable Posterior Distributions
• Divergent Effects of Factors on Crashes under Autonomous and Conventional Driving Modes Using A Hierarchical Bayesian Approach
• Rao-Blackwellization in the MCMC era
• Model-based metrics: Sample-efficient estimates of predictive model subpopulation performance
• Advances in Importance Sampling
• Variational Bayes latent class approach for EHR-based phenotyping with large real-world data
• Region-Referenced Spectral Power Dynamics of EEG Signals: A Hierarchical Modeling Approach
• q-Paths: Generalizing the Geometric Annealing Path using Power Means
• Bayesian model selection in the $\mathcal{M}$-open setting -- Approximate posterior inference and probability-proportional-to-size subsampling for efficient large-scale leave-one-out cross-validation
• Conformal Bayesian Computation