| Frequently analysts use simulation to compare several alternatives at various experimental points. For example, analysts may use simulation to compare several scheduling rules in a time-shared computer system at various experimental points. Moreover, these experimental points may be drawn from a high-dimensional experimental space (i.e., an experimental space composed of many experimental factors, such as job arrival rate, expected job size, number of various types of service facilities, buffer sizes in front of various service facilities, etc.).;Basically, the focus of the simulation comparison is to find out which alternative performs well on which experimental points. However, decision makers may ask: How well does an alternative perform overall? Some research questions then follow: Can we measure the performance of an alternative in an overall way? And if so, how? Answering questions such as these is the focus of this research.;This research defines a measure of the overall performance, illustrates its usefulness, and recommends practical guidelines for its use. |
