Using Machine Learning for Heterogeneous Treatment Effect Estimation

Epidemiological Tools Seminar
MH-1407

Epi Tools Workshop: Susan Athey, PhD, Professor of Economics, Stanford University Graduate School of Business

Machine learning can be used not just for prediction, but also for estimating the causal effects of treatments.  Valid confidence intervals can also be obtained when machine learning methods are appropriately modified. This seminar will outline methods for estimating heterogeneous treatment effects and personalized treatment assignment policies in randomized experiments and observational studies, and will illustrate the methods with several applications.

 

Using Machine Learning for Heterogeneous Treatment Effect Estimation