This paper introduces a new approach to estimating the propensity score using Gaussian processes and optimizing hyperparameters with respect to covariate balance.
Sep 19, 2019
Brian Vegetabile is an associate statistician at the RAND Corporation. His prior research was motivated by problems arising from the Conte Center on Brain Programming and Adolescent Vulnerabilities at the University of California, where he also studied methods for quantifying predictability in mother-infant interactions as a marker for later behavioral outcomes in observational data settings. His current methodological research interests are in Bayesian inference and statistical machine learning techniques as they apply to causal inference in observational studies. Vegetabile earned a B.S. in aerospace engineering from Penn State University and spent four years working as a satellite systems engineer for Northrop Grumman. His Ph.D. thesis focused on methods for obtaining optimal covariate balance for causal inference in observational studies. He obtained his Ph.D. in statistics from the University of California, Irvine.