时间🤰：2019年6月11日（周二）下午1:00-3:30
地点🧎🏻♀️➡️：中北校区理科大楼A1716报告厅
时间🗄🚣♂️：2019年6月12日（周三）上午8:30-11:00
地点：中北校区理科大楼A1716报告厅
题目🫃🏿：From Randomization to Covariate Balancing: Applications of Propensity Score Methods to Causal Inference
主讲人：Professor Menggang Yu（Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison）
摘要：
A central goal of most scientific inquiries is to infer‘cause and effect’ that describes what would have happened if the environment was changed through certain interventions. This short course will provide an introduction to statistical theory of causal inference, and cast the study of causation in a formal statistical framework. Due to the link between causal inference and missing data, propensity score and related methods will be the main techniques for our inference in this course. A variety of propensity score based methods will be connected to the fundamental idea of randomization (including stratified randomization and post-randomization stratification). Topics with high dimensional covariates and effective study designs will also be covered.
Tentative topics:
1.Randomization in clinical trials
2.Observational studies, potential outcomes, and the Rubin causal model
3.Propensity scores for causal inference
4.Double robust estimation and semiparametric efficiency
5.Covariate balancing propensity score (CBPS)
6.Variations on CBPS and entropy balancing
7.Propensity score with high-dimensional covariates
8.Utility of propensity score in study designs