The Department of Statistics and Actuarial Science Fall Colloquium Series presents:

Yuling Yao, Assistant Professor, Department of Statistics and Data Sciences, University of Texas at Austin

"Post-stratified importance sampling: I cannot believe you have not used MRP in importance weighting."

Abstract: In this talk, I will introduce “Post-stratified importance sampling.” Importance sampling is the building block of modern statistical computing and modeling, but the weighted estimate is prone to high or infinite variance. Moreover, its empirical standard error is often unreliable, leaving importance sampling without practical uncertainty quantification. Inspired by the heuristics of multilevel regression and poststratification (MRP) in survey sampling, we propose post-stratified importance sampling (PIS), applicable to the generic setting of computing expectations using proposal draws. PIS turns an arbitrarily high-dimensional weighting task into two one-dimensional probabilistic modeling problems: regression for the conditional mean given the importance ratio, and marginal density estimation for the importance ratio. Different modeling choices yield a family of PIS estimators, encompassing ordinary importance sampling as a special case. We develop a model-based implementation that combines one-dimensional Gaussian process regression for the conditional mean and score-informed heavy-tailed density estimation for the marginal law, using both the observed weights and the available density evaluations. At a higher level, the development of PIS is an example to demonstrate the benefit of bringing the modeling workflow into statistical computing.

This virtual presentation begins at 3:20 p.m.

Topic: Colloquium 2026-2027

Time: Nov. 5, 2026 3:20 p.m. Central Time (US and Canada)

Join Zoom Meeting

Meeting ID: 915 1899 4105