Report Title: Causal Inference Machine Learning for Dynamic Decision Making

Speaker: Professor Wang Lu

Affiliation: Department of Biostatistics, University of Michigan

Report Time: August 2, 2023 (Wednesday) 10:00-11:00 AM

Report Location: Room 1602, Block A, Emerald Science and Education Building

Report Abstract: In this talk, we present recent advances and statistical causal learning developments for evaluating Dynamic Treatment Regimes (DTR), which allow the treatment to be dynamically tailored according to evolving subject-level data. Identification of an optimal DTR is a key component for precision medicine and personalized health care. We will first present a tree-based doubly robust reinforcement learning (T-RL) method, which builds a decision tree that maintains the nature of batch-mode reinforcement learning, and then a new Stochastic-Tree Search method called ST-RL for evaluating optimal DTRs, which contributes to the existing literature in its non-greedy policy search and demonstrates outstanding performances even with a large number of covariates. In addition, we consider a common challenge with practical “restrictions” and develop a Restricted Tree-based Reinforcement Learning (RT-RL) method to address this challenge. We illustrate the method using an observational dataset to estimate a two-stage stepped-up DTR for guiding the level of care placement for adolescents with substance use disorder.

Personal Introduction: Wang Lu, Ph.D., is currently a tenured professor and deputy director of the Department of Biostatistics at the University of Michigan. She graduated from Peking University in 2002 and received her Ph.D. from Harvard University in 2008. Her research areas include statistical methods for evaluating and optimizing dynamic treatment plans, personalized medicine, causal inference, non-parametric and semi-parametric regression, missing data analysis, and longitudinal (related/clustered) data analysis. She has published more than 139 papers in academic journals such as JASA, Biometrika, Biometrics, AoAS, and co-authored a book chapter. She is currently the associate editor of JASA and Biometrics.