Large-deviation sampling laws for constrained simulation optimization on finite sets

We consider the problem of selecting an optimal system from among a finite set of competing systems, based on a “stochastic” objective function and subject to a single “stochastic” constraint. By strategically dividing the competing systems, we derive a large deviations sampling framework that asymptotically minimizes the probability of false selection. We provide an illustrative example where a closed-form sampling law is obtained after relaxation.