Analysis of simulation output using sequential batch means

ANALYSIS OF SIMULATION OUTPUT USING SEQUENTIAL BATCH MEANS METHOD CM. Sc. Thesis} Fulya ALTIPARMAK GAZI UNIVERSITY INSTITUTE OF SCIENCE AND TECHNOLOGY September - 1990 ABSTRACT The purpose of statistical output analysis is to predict the performans of a system, or to compare the performance of two or more alternative system desing. Therefore, statistical analysis of the output is an essentially part of a simulation study. Constructing the confidence interval however, is difficult because the output data from a simulation are in general nonstationary and autocorrelated, so that direct application of the techniques of classical statistics is precluded. There are several methods Cfixed sample size methods and sequential determining the length of a simulation methods!) for constructing confidence interval in the simulation literature. In this study, in order to construct a 10QCl-oû% confidence interval for steady state mean, which is called p, it has been developed of Adam's algorithm by using sequential batch means' procedure. İV