Robust design for the estimation of a threshold probability

Abstract

We consider the construction of robust sampling designs for the estimation of threshold probabilities in spatial studies. A threshold probability is a probability that the value of a stochastic process at a particular location exceeds a given threshold. We propose designs such that the estimation of threshold probabilities is robust to two possible model uncertainties: misspecified regression responses and covariance structures. To address these two uncertainties of this stochastic process, we average the mean squared error of the predicted values relative to the true values, over all possible covariance structures in a neighbourhood of the experimenter's nominal choice, and then maximize it over a neighbourhood of the fitted model. Finally, the maximum is minimized to obtain the robust designs.