A modification of Cheng'S method: An alternative Factor Screening method for stochastic simulation models

Abstract

Factor Screening experiments identify those factors with significant effect on a selected output. We propose a modification of Cheng's method as a new factor screening alternative for simulation models whose output has homogeneous variance and can be described by a second-order polynomial function. The performance of the proposed model is compared with several other factor screening alternatives through an empirical evaluation. The results show that the proposed method sustains its efficiency and accuracy as the number of factors or the homogeneous variance increases. However, its accuracy degrades as variance heterogeneity increases.