Join Advance-CTR and the Data Science Initiative at Brown for this 5-part series exploring machine learning, its methodology, and application in biomedicine and health. The purpose of this series is to serve as an introduction to machine learning for researchers, clinician scientists, and others who may be interested in using these methods in their research.
Wednesday, October 13, 2021:
Yichi Zhang, PhD: “Interpretable Individualized Treatment Rules Using Decision Lists”
Precision medicine is currently a topic of great interest in clinical science. One typical way to formalize precision medicine is through an individualized treatment rule, which is a sequence of rules, one per each stage of intervention, that map up-to-date patient information to a recommended treatment. An optimal individualized treatment rule is defined as maximizing the mean of some cumulative clinical outcome if applied to a population of interest. In many settings, estimation of an optimal individualized treatment rule is an exploratory analysis intended to generate new hypotheses for subsequent research and not to directly dictate treatment to new patients. In such settings, a rule that is interpretable in a domain context may be of greater value than an unintelligible one built using “black-box” methods. In this talk, I will present a causal inference framework for estimating an optimal individualized treatment rule and discuss its connection to reinforcement learning. Then, I will describe an estimator of an optimal and interpretable rule, which is expressible as a list of “if-then” statements that can be presented as either a paragraph or as a simple flowchart that is immediately interpretable to domain experts. The proposed method will be illustrated using a clinical trial dataset.
About the Speaker
Yichi Zhang is an assistant professor in the Department of Computer Science and Statistics at the University of Rhode Island. He received his PhD in Statistics from North Carolina State University and received postdoctoral training at Harvard School of Public Health. His research focuses on data-driven decision-making and sequential causal inference with applications to biomedical problems.