TABLE OF CONTENTS Correlation analysis is a modeling and analysis approach that straddles input analysis and output analysis. It consists of two activities:
Modeling correlated stochastic processes. Generally, both autocorrelations within individual time series and cross-correlations across multiple time series are of interest (see Section 3.6). However, modeling autocorrelations preponderates
Studying the impact of correlations on performance measures of interest via sensitivity analysis. In such studies, the modeler may "increase" or "decrease" the magnitude of correlations in various correlated processes, and then observe the resulting performance measures (see Section 9.7). If the impact is marginal, modeling correlations can be dispensed with
Thus, correlation analysis combines modeling (the input analysis of the first item) with analysis (the output analysis in the second item), especially in the context of simulation. As discussed later, correlation analysis is motivated by the impact of correlations on performance measures and the fact that ignoring them can lead to large errors in predicting system performance. Applications to manufacturing systems will be discussed later.
Correlation analysis as part of input analysis is simply an approach to modeling and data fitting. This approach insists on high-quality models incorporating temporal dependence, and strives to fit correlation-related statistics in a systematic way. To set the scene, consider a stationary time series {X n} ? n = 0, that is, a time series in which all statistics remain unchanged under the passage of time. In particular, all X n share a common mean, m
