Maximum One-Factor-At-A-Time Designs for Screening in Computer Experiments

Identifying important factors from a large number of potentially highly nonlinear and computationally expensive black box model is difficult problem. Morris screening Sobol’ design are two commonly used model-free methods for doing this. In this article, we establish connection between these seemingly different in terms their underlying experimental structure further exploit to develop an improved called Maximum One-Factor-At-A-Time (MOFAT) design. We also efficient constructing MOFAT designs with factors. Several examples presented demonstrate the advantages compared methods.