Rotated Factorial Designs for Computer Experiments

Scott D. Beattie ； Dennis K. J. Lin

《中國統計學報》 42卷4期 (2004/12) Pp. 431-450

https://doi.org/10.29973/JCSA.200412.0007

並列摘要

Computer models can describe complicated phenomena encountered in science and engineering fields. To use these models for scientific investigation, however, their generally long running times and mostly deterministic nature require a specially designed experiment. This paper presents a class of Latin hypercube designs developed from the rotation of factorial designs. These rotated factorial designs are easy to construct and preserve many of the attractive properties of standard factorial designs: they have equally-spaced projections to univariate dimensions and uncorrelated regression effect estimates (orthogonality). A thorough application of the newly proposed design in mechanical engineering is presented. New results on the rotated mixed-level factorial designs are discussed.

並列關鍵字

Effect correlation ； minimum interpoint distance ； mixed level ； rotated factorial design

參考文獻

Beattie, S. D.(1999).Contributions To the Design and Analysis of Experiments.The Pennsylvania State University, Department of Statistics.

Beattie, S. D.,D. K. J. Lin(1998).Rotated factorial designs for computer experiments.Technical Report TR#98-02.(Technical Report TR#98-02).,未出版The Pennsylvania State University, Department of Statistics.

Beattie, S. D.,D. K. J. Lin,J. Q. Fan,G. Li, forthcoming(ed.)(2005).A New Class of Latin Hypercube for Computer Experiments.(Chapter in Contemporary Multivariate Analysis and Experimental Designs).

Bursztyn, D.,D. M. Steinberg(2001).Rotation designs for experiments in high bias situations.Journal of Statistical Planning and Inference.97,399-414.

Bursztyn, D.,D. M. Steinberg(2002).Rotation designs: Orthogonal first-order designs with higher order projectivity.Journal of Applied Stochastic Models in Business and Industry.18,197-206.