ABSTRACT
In this chapter, the authors discuss how the great control the analyst has over the model can be used to optimize the system, and how the simulation process can be used to solve complex mathematical and statistical problems that defy analytical solution. Simulation is a powerful method not only for studying and optimizing real-world systems, but also for solving complex mathematical and statistical problems that defy analytical solution. In recent years, what is called the “bootstrap method” has gained popularity in both statistical and non-statistical circles for the solution of such problems. The author summarizes a recent improvement, called the Generalized Bootstrap, which can avoid the drawbacks of the naive bootstrap. When a real-world system under simulation may be operated in two ways, a goal often of interest to the experimenter is to quantify the difference in a key output measure’s mean value under the two modes of operation.
