Published January 1, 2018 | Version v1
Journal article Open

Estimates of the coverage of parameter space by Latin Hypercube and Orthogonal Array-based sampling

  • 1. Univ Queensland, Sch Math & Phys, Brisbane, Qld 4072, Australia
  • 2. Queensland Univ Technol, ARC Ctr Excellence Math & Stat Frontiers, Brisbane, Qld, Australia
  • 3. Univ Plymouth, Dept Math & Stat, Plymouth, Devon, England
  • 4. Koc Univ, Dept Math, TR-34450 Istanbul, Turkey

Description

In this paper we use counting arguments to prove that the expected percentage coverage of a d dimensional parameter space of size n when performing k trials with either Latin Hypercube sampling or Orthogonal Array-based Latin Hypercube sampling is the same. We then extend these results to an experimental design setting by projecting onto a t < d dimensional subspace. These results are confirmed by simulations. The theory presented has both theoretical and practical significance in modelling and simulation science when sampling over high dimensional spaces. (C) 2017 Elsevier Inc. All rights reserved.

Files

bib-3880d70d-209f-4e61-b140-565cc8741e5f.txt

Files (237 Bytes)

Name Size Download all
md5:de37ed89f29c850de212d6d47c85ca18
237 Bytes Preview Download