AG Kommunikationstheorie

Thema:

Variance estimation in quantitative simulation

Abstract:

In the output data analysis of steady-state simulation experiments, it is desired to find estimates of the process variance parameter. With its help it is possible to estimate the mean square error in the estimation of the process mean, and give valid confidence intervals for the process mean. Finally, in comparing performance of two systems under investigation, the variance parameter itself may be of interest.

This talk will introduce the process variance parameter, present the Standardized Time Series Area Estimator for its estimation, and finally give some results from the implementation of the estimator.



zurück zur Terminübersicht