Number‐Theoretic Methods

Fang, Kai‐Tai; Wang, Yuan

doi:10.1002/0471667196.ess0697

Cited by 1 publication

(1 citation statement)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Wand and Wang (2005) advanced two‐dimensional sequential uniform design by integrating the thoughts of sequence into uniform design (Fang, 1994; Wang and Fang, 1996; Fang and Ma, 2002; Wang et al , 2006) and they also resolved the problem of the validity of revolving monotone function. Later, Huang and Peng (2005) generalized Wang and Wang's idea and provided two‐dimensional sequential β, c design, which resolved the validity of revolving monotone function.…”

Section: Introductionmentioning

confidence: 99%

Three‐dimensional sequential uniform design

2009

View full text Add to dashboard Cite

Purpose -The purpose of this paper is to extend two-dimensional sequential uniform design and present three-dimensional sequential uniform design which can optimize the problem of determining the extreme value of three-factor polynomial. Design/methodology/approach -Since there are limitations in two-dimensional sequential uniform design, three-dimensional sequential uniform design introduced in this paper is to arrange the experimental points by U 9 ð9 3 Þ uniform design table and to determine the maximum of the experimental values according to the demand. It is proven that the convergence of the experimental central point's sequence and that it can optimize the problem of determining the extreme value of no-cross term quadratic polynomial monotone function class and the corresponding conditions are also proven. Taking the no-cross term quadratic polynomial monotone function as an example, the superiority of it is testified. Findings -Three-dimensional sequential uniform design can optimize the problem of determining the extreme value of three-factor polynomial. It can get higher precision and better convergence but do fewer experiments than general uniform design. Practical implications -A very effective method in resolving the problem of selecting optimum on multi-dimensional space. Originality/value -A new method of three-dimensional sequential uniform design which can get higher precision and better convergence is presented.

show abstract

Section: Introductionmentioning

confidence: 99%