Introduction to Stochastic Simulation Optimization

“Stochastic simulation optimization” (often shortened as simulation optimization) refers to stochastic optimization using simulation. Specifically, the underlying problem is stochastic and the goal is to find the values of controllable parameters (decision variables) to optimize some performance measures of interest, which are evaluated via stochastic simulation, such as discrete-event simulation or Monte Carlo simulation [Fu, 2002; Chen et al., 2008; Fu et al., 2008]. The act of obtaining the best decision under given circumstances is known as optimization. It is one of the leading topics of study and research. Optimization problems can be broadly classified into several branches depending upon the nature of the problem. For deterministic optimization problems, no randomness or uncertainty is considered and so the solutions have been relatively simple. The deterministic optimization approach might work well when the real problem has no noise or the uncertainty is not critical. However, many real-world optimization problems involve some sort of uncertainties in the form of randomness. These problems are studied under the branch “stochastic optimization” which plays a significant role in the design, analysis, and operation of modern systems. Methods for stochastic optimization provide a means of coping with inherent