Machine Learning Seminar Series

Efficient Discovery of Heterogeneous Treatment Effects in Randomized Experiments via Anomalous Pattern Detection


Edward McFowland III
Information & Decision Sciences
Carlson School of Management
University of Minnesota

Wednesday, November 11, 2020
3:30–4:30 pm

Online via zoom view recording here

In the recent literature on estimating heterogeneous treatment effects, each proposed method makes its own set of restrictive assumptions about the intervention's effects and which subpopulations to explicitly estimate. Moreover, the majority of the literature provides no mechanism to identify which subpopulations are the most affected--beyond manual inspection—and provides little guarantee on the correctness of the identified subpopulations. Therefore, we propose Treatment Effect Subset Scan (TESS), a new method for discovering which subpopulation in a randomized experiment is most significantly affected by a treatment. We frame this challenge as a pattern detection problem where we efficiently maximize a nonparametric scan statistic over subpopulations. Furthermore, we identify the subpopulation which experiences the largest distributional change as a result of the intervention, while making minimal assumptions about the intervention's effects or the underlying data generating process. In addition to the algorithm, we demonstrate that the asymptotic Type I and II error can be controlled, and provide sufficient conditions for detection consistency—i.e., exact identification of the affected subpopulation. Finally, we validate the efficacy of the method by discovering heterogeneous treatment effects in simulations and in real-world data from a well-known program evaluation study.

Dr. Edward McFowland III is an Assistant Professor of Information and Decision Sciences in the Carlson School of Management, at the University of Minnesota; he received his Ph.D. in Information Systems and Management from Carnegie Mellon University. Dr. McFowland’s research interests—which lie at the intersection of Information Systems, Machine Learning, and Public Policy—include the development of computationally efficient algorithms for large-scale statistical machine learning and “big data” analytics. More specifically, his research seeks to demonstrate that many real-world problems faced by organizations, and society more broadly, can be reduced to the tasks of anomalous pattern detection and discovery. As a data and computational social scientist, Dr. McFowland’s broad research goal is bridging the gap between machine learning and the social sciences (e.g., economics, public policy, and management) both through the application of machine learning methods to social science problems and through the integration of machine learning and econometric methodologies.

Dr. McFowland’s research has been supported by Adobe, Facebook, PNC Bank, AT&T Research Labs, and the National Science Foundation; his work has been published in leading Management, Machine Learning, and Statistics journals and conferences.