You are cordially invited to the seminars organized by the Department of Mathematics.<\/p>\n
Speaker: G\u00f6zde Sert (Texas A&M University)<\/p>\n
\u201cBayesian Semiparametric Causal Inference: Targeted Doubly Robust Estimation of Treatment Effects\u201d<\/p>\n
Abstract: We propose a semiparametric Bayesian methodology for estimating the average treatment effect (ATE) within the potential outcomes framework using observational data with high-dimensional nuisance parameters. Our approach introduces a Bayesian debiasing procedure that corrects for bias arising from nuisance estimation and employs a targeted modeling strategy based on summary statistics rather than the full data. These summary statistics are identified in a debiased manner, enabling estimation of nuisance bias via weighted observables and facilitating hierarchical learning of the ATE. By combining debiasing with sample splitting, the proposed method separates nuisance estimation from inference on the target parameter, thereby reducing sensitivity to nuisance model misspecification. We establish that the marginal posterior for the ATE satisfies a Bernstein\u2013von Mises theorem when both nuisance models are correctly specified, and remains consistent and robust when only one is correctly specified, achieving Bayesian double robustness. Extensive simulations confirm the theoretical results, demonstrating accurate point estimation and credible intervals with nominal coverage, even in high-dimensional settings.<\/p>\n
Date: February 9, Monday
\nTime: 19:00 (Ankara)
\nPlace: Zoom<\/p>\n
To request the event link, please send a message to gokhan.yildirim@bilkent.edu.tr<\/p>\n","protected":false},"excerpt":{"rendered":"
You are cordially invited to the seminars organized by the Department of Mathematics. Speaker: G\u00f6zde Sert (Texas A&M University) \u201cBayesian Semiparametric Causal Inference: Targeted Doubly Robust Estimation of Treatment Effects\u201d Abstract: We propose a semiparametric Bayesian methodology for estimating the average treatment effect (ATE) within the potential outcomes framework using [\u2026] <\/p>\n