All Times EDT
Keywords: targeted estimation, stochastic regime, super learning
Targeted estimation method proposed by van dee laan has many attractive properties including producing unbiasedness, efficiency, and well-defined estimator. This makes it preferable to other existing methods, especially in real-world studies including pragmatic clinical trials. Stochastic treatment effects correspond to stochastic regimes in which the assigned treatment is allowed to depend on the natural value of treatment, which can therefore be defined to be more realistic and relevant than deterministic regimes. This talk will discuss the targeted estimation method to define, identify, and estimate treatment effects of this type of regime.