A non-parametric comparison algorithm for simulation optimization

In this paper, we present a non-parametric sequential search algorithm for solving discrete stochastic optimization problem where the objective function does not have analytical form, but has to be measured or estimated, for instance through Monte Carlo simulation. The optimization algorithm we present in this paper uses binary hypothesis test. At each iteration of the algorithm, two neighboring configurations are compared and the one that appears to be better is passed on to the next iteration. The algorithm uses a sequential sampling procedure with increasing boundaries as the number of iteration increases. We show that under suitable conditions on the boundaries, the algorithm converges almost surly to an optimum solution. Computational results will be given to demonstrate the performance of the proposed approach.