Haidong Li

Near-Oracle Robustification of Finite-Difference Stochastic Gradient Estimators via Cheap Pilot Calibration

It is shown that, by pilot-estimating these model quantities using a negligible fraction of the simulation budget, substantial robustness is attained in the resulting FD estimators, and how such an approach is competitive against any choices of prescribed perturbation size, even if they are designed to be minimax-optimal over reasonable classes of target functions and FD schemes.