Optimal maximin $L_{1}$-distance Latin hypercube designs based on good lattice point designs

@article{Wang2018OptimalM,
  title={Optimal maximin \$L\_\{1\}\$-distance Latin hypercube designs based on good lattice point designs},
  author={Lin Wang and Qian Xiao and Hongquan Xu},
  journal={The Annals of Statistics},
  year={2018},
  url={https://api.semanticscholar.org/CorpusID:13671361}
}
This work constructs a series of maximin Latin hypercube designs via Williams transformations of good lattice point designs that are optimal under the maximin L1-distance criterion, while others are asymptotically optimal.
View via Publisher
projecteuclid.org
45 Citations
Highly Influential Citations
4
Background Citations
13
Methods Citations
11

Figures and Tables from this paper

45 Citations

On maximin distance and nearly orthogonal Latin hypercube designs

Optimal maximin L2-distance Latin hypercube designs

Construction of Flexible Maximin Latin Hypercube Designs Based on Good Lattice Point Sets

This paper would like to further this idea for more general, or more flexible, designs, such as N equal to primes, prime multiples and prime powers, and implement a similar construction algorithm to build optimal or asymptotically optimal designs of such dimension.

A new non-iterative deterministic algorithm for constructing asymptotically orthogonal maximin distance Latin hypercube designs

A new non-iterative deterministic algorithm for constructing asymptotically orthogonal maximin distance LHDs and an iterative algorithm utilizing a mixture of criteria is provided for further improvement of the performance of the newly constructed L HDs from multiple perspectives.

Musings about Constructions of Efficient Latin Hypercube Designs with Flexible Run-sizes

Using the R package which integrates and improves various algebraic and searching methods, many of the designs found in this paper are better than the existing ones and can serve as benchmarks for the future developments on LHDs.

A Construction Method for Maximin L1-Distance Latin Hypercube Designs

A Construction Method for Maximin L1-Distance Latin Hypercube Designs

LHD: An R package for efficient Latin hypercube designs with flexible sizes

The R package LHD is developed that integrates and improves various algebraic and searching methods to construct efficient LHDs with flexible design sizes and many of the designs found in this paper are better than the existing ones.

A method of constructing maximin distance designs

The basic idea is to construct large designs from small designs and the method is effective because the quality of large designs is guaranteed by that of small designs, as evaluated by the maximin distance criterion.

New bounds and search for maximin distance U‐type designs

This study obtains a new upper bound for the separation distance of certain classes of five‐level U‐type designs, investigates the characteristics of maximin distance U‐type designs and shows the optimality of some existing orthogonal designs.

Connecting U-type Designs Before and After Level Permutations and Expansions

This paper systematically study the theoretical properties of U-type designs before and after level permutations and expansions, and establishes the relationships between the initial designs’ generalized word-length patterns (GWLP) and the generated designs' orthogonal and space-filling properties.

33 References

Construction of maximin distance Latin squares and related Latin hypercube designs

Based on number theory and finite fields, three algebraic methods to construct maximin distance Latin squares as special Latin hypercube designs are proposed and lower bounds on their minimum distances are developed.

ORTHOGONAL-MAXIMIN LATIN HYPERCUBE DESIGNS

A multi-objective optimization approach to find good LHDs by combining correlation and distance performance measures is proposed and a new exchange algorithm for efficiently generating such designs is proposed.

Algorithms for Generating Maximin Latin Hypercube and Orthogonal Designs

A new, well-performing algorithm is presented for the construction of maximin Latin hypercube designs using a 2-dimensional distance metric and an additional criterion, design orthogonality, is important when screening the effects of the input variables and a new search algorithm for orthogonal maximin designs is described.

Construction of orthogonal Latin hypercube designs

Latin hypercube designs have found wide application. Such designs guarantee uniform samples for the marginal distribution of each input variable. We propose a method for constructing orthogonal Latin…

Optimal and orthogonal Latin hypercube designs for computer experiments

Latin hypercube designs suitable for factor screening are presented and they are shown to be efficient in terms of runs required per factor as well as having optimal and orthogonal properties.

A construction method for orthogonal Latin hypercube designs

The method generates orthogonal Latin hypercube designs that can include many more factors than those proposed by Ye (1998) and can also be used to construct Latin hyper cube designs with low correlation of first-order and second-order terms.

Optimal Sliced Latin Hypercube Designs

A new construction approach is developed that first generates the small Latin hypercube design in each slice and then arranges them together to form the SLHD.

Efficient Nearly Orthogonal and Space-Filling Latin Hypercubes

An algorithm for constructing orthogonal Latin hypercubes, given a fixed sample size, in more dimensions than previous approaches is presented, and a method that dramatically improves the space-filling properties of the resultant Latinhypercubes is detailed.

Orthogonal Array-Based Latin Hypercubes

It is proved that when used for integration, the sampling scheme with OA-based Latin hypercubes offers a substantial improvement over Latin hypercube sampling.

CONSTRUCTION OF ORTHOGONAL AND NEARLY ORTHOGONAL LATIN HYPERCUBE DESIGNS FROM ORTHOGONAL DESIGNS

An approach to constructing 2 r -order orthogonal LHDs with 2 r+1 +2 runs and 2 r factors has the minimum correlation between any two distinct columns.