Alternative Foldover Plans for Two-Level Nonregular Designs

作者:Edwards David J*; Brooks J Paul
来源:Communications in Statistics - Simulation and Computation, 2014, 43(1): 209-224.
DOI:10.1080/03610918.2012.700361

摘要

Foldover is a classic technique used to select follow-up experimental runs when an initial experiment yields ambiguities. While foldover has been soundly investigated for regular designs, less research has been devoted to this technique for nonregular designs. Previous work focuses on the use of the generalized minimum aberration criterion to obtain optimal foldover plans. In contrast, this article utilizes the concept of minimal dependent sets (MDSs) and associated criteria to rank foldovers of nonregular designs. We propose an integer programming-based solution to aid in the location and enumeration of MDSs. MDS-optimal foldovers for selected nonregular designs are presented and discussed.

  • 出版日期2014-1-1

全文