Online Program

Targeted Machine Learning for Causal Inference with Real-World Data (301230)

*Rachael Phillips, University of California, Berkeley 
Mark van der Laan, University of California, Berkeley 

Keywords: targeted learning, super learning, machine learning, real-world evidence

In this talk, we introduce a principled and reproducible approach, termed Targeted Learning, for estimating target estimands of interest from complex, real-world data. Targeted estimators are defined within the context of realistic semiparametric models, and are multiply robust, efficient plug-in estimators. State-of-the-art machine learning is incorporated in the estimation strategy to flexibly adjust for confounding while yielding valid statistical inference. We present a roadmap for Targeted Learning and highlight real-world applications, from longitudinal data with intercurrent events to randomized control trials. We will also discuss the utility of this robust estimation strategy in comparison to conventional techniques.