Presented by Dr. Susan Gruber, biostatistician, and founder of Putnam Data Sciences, LLC. Targeted maximum likelihood estimation (TMLE) is an efficient, double robust procedure for learning from data. TMLE can be used to appropriately adjust for selection bias, informative loss to follow-up, and other intercurrent events that threaten validity of causal effect estimates. General principles will be introduced via a concrete example, showing how TMLE can be used to estimate the marginal additive effect of treatment versus comparator.
