Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
7548830 | Statistics & Probability Letters | 2018 | 7 Pages |
Abstract
We propose a new construction for nested designs, called General Nested Latin Hypercube designs (GNLHs). Such designs contain nested Latin hypercube designs as special cases. Besides achieving maximum uniformity in one dimension, each layer of GNLHs is flexible in run sizes. Moreover, theoretical results and numerical simulations show that GNLHs perform well on the sampling variance.
Related Topics
Physical Sciences and Engineering
Mathematics
Statistics and Probability
Authors
Jin Xu, Xiaojun Duan, Zhengming Wang, Liang Yan,