Doubly-robust estimation of shared-parameter adaptive treatment strategies
Friday, September 15, 2017 - 2:30pm - 3:00pm
In this talk, I will introduce doubly-robust G-estimation of an adaptive treatment strategy in which parameters are shared across different stages of the treatment sequence, allowing for more efficient estimation and simpler treatment decision rules. The approach is computationally stable, and produces consistent estimators provided either the outcome model or the treatment allocation model is correctly specified. An illustrative example is also provided.