Targeted Learning for Estimating Causal Effects in Observational and Randomized Studies
Speaker: Susan Gruber, PhD, Sr. Director of IMEDS Methods Research,Reagan-Udall Foundation for the Food & Drug Admin
Targeted Learning offers a principled statistical approach to answering questions about health and safety from data. The targeted learning framework combines two state of the art methodologies, super learning (SL) for data adaptive machine learning, and targeted minimum loss-based estimation (TMLE) for efficient semi-parametric estimation. The presentation introduces the Targeted Learning estimation roadmap, and describes TMLE and Super Learning for estimating causal effects.