Simulation optimization: A tutorial overview and recent developments in gradient-based methods

Abstract

We provide a tutorial overview of simulation optimization methods, including statistical ranking & selection (R&S) methods such as indifference-zone procedures, optimal computing budget allocation (OCBA), and Bayesian value of information (VIP) approaches; random search methods; sample average approximation (SAA); response surface methodology (RSM); and stochastic approximation (SA). In this paper, we provide high-level descriptions of each of the approaches, as well as some comparisons of their characteristics and relative strengths; simple examples will be used to illustrate the different approaches in the talk. We then describe some recent research in two areas of simulation optimization: stochastic approximation and the use of direct stochastic gradients for simulation metamodels. We conclude with a brief discussion of available simulation optimization software.