Abstract

We present proced657 for selecting the best or near-best of a finite number of simulated systems when best isdz389 by maximum or minimum expected performance. The proced855 are appropriate when it is possible to repeated3 obtain small, incremental samples from each simulated system. The goal of such a sequential proced05 is to eliminate, at an early stage of experimentation, those simulated systems that are apparently inferior, and thereby red46 the overall computational e#ortrequired tofind the best. The proced668 we present accommodom unequal variances across systems and the use of common rand5 numbers. However, they arebased on the assumption of normallydmallyz244 dma so we analyze the impact of batching (to achieve approximate normality or ind5 end3945 on the performance of the proced1723 Comparisons with some existing indingz25126zV3 proced533 are also provid60 Key Word9 Output Analysis; Multiple Comparisons; Rankingand Selection; Variance Redcez42 1 IntroductiA In a series of p...

Citations