A Methodology for Factor Screening for Multiple Response Computer Simulation Models

The majority of the literature dealing with simulation output analysis has focused on single response models with a limited number of input variables. However, most real world simulation applications involve multiple performance measures and numerous input variables. The purpose of this session is to outline a methodology for factor screening in the presence of multiple simulation responses. The intent is to provide general guidelines to reducing the original set of n input factors to a subset of k factors which exert the most significant impact on a set of m performance measures.