Stochastic search and optimization in discrete event systems: an overview of parametric and nonparametric methods

We provide an introductory overview of stochastic search algorithms applied to optimization problems in discrete event systems. We classify such optimization problems into two categories: parametric and nonparametric. The optimization of discrete event systems typically relies on estimates of performance related quantities via simulation or online observation. Therefore, the use of stochastic search algorithms must account for estimation errors and the potential complexity of implementing such estimators.