Simulation-based optimization using simulated annealing with ranking and selection

In this paper, we present a new iterative method that combines the simulated annealing method and the ranking and selection procedures for solving discrete stochastic optimization problems. The number of visits to every state by the proposed algorithm is used to estimate the optimal solution. We show that the configuration that has been visited most often in the first m iterations converges almost surely to a globally optimum solution. We present empirical results that illustrate the performance of the proposed method.