Maximin Latin Hypercube Designs in Two Dimensions

The problem of finding a maximin Latin hypercube design in two dimensions can be described as positioning n nonattacking rooks on an n A— n chessboard such that the minimal distance between pairs of rooks is maximized. Maximin Latin hypercube designs are important for the approximation and optimization of black-box functions. In this paper, general formulas are derived for maximin Latin hypercube designs for general n, when the distance measure is l∞ or l1. Furthermore, for the distance measure l2, we obtain maximin Latin hypercube designs for n ≤ 70 and approximate maximin Latin hypercube designs for other values of n. All these maximin Latin hypercube designs can be downloaded from the website http://www.spacefillingdesigns.nl. We show that the reduction in the maximin distance caused by imposing the Latin hypercube design structure is small. This justifies the use of maximin Latin hypercube designs instead of unrestricted designs.

[1]  M. E. Johnson,et al.  Minimax and maximin distance designs , 1990 .

[2]  Rex K. Kincaid,et al.  Approximate Solutions of Continuous Dispersion Problems , 2005, Ann. Oper. Res..

[3]  Dick den Hertog,et al.  Optimizing color picture tubes by high-cost nonlinear programming , 2002, Eur. J. Oper. Res..

[4]  Ken R. McNaught,et al.  A comparison of experimental designs in the development of a neural network simulation metamodel , 2004, Simul. Model. Pract. Theory.

[5]  Janis Auzins,et al.  Response surface method for solution of structural identification problems , 2004 .

[6]  Marco Locatelli,et al.  Packing equal circles in a square: a deterministic global optimization approach , 2002, Discret. Appl. Math..

[7]  A. J. Booker,et al.  A rigorous framework for optimization of expensive functions by surrogates , 1998 .

[8]  Dick den Hertog,et al.  Constrained Maximin Designs for Computer Experiments , 2003, Technometrics.

[9]  Tibor Csendes,et al.  A New Verified Optimization Technique for the "Packing Circles in a Unit Square" Problems , 2005, SIAM J. Optim..

[10]  Patric R. J. Östergård,et al.  More Optimal Packings of Equal Circles in a Square , 1999, Discret. Comput. Geom..

[11]  Jerome Sacks,et al.  Designs for Computer Experiments , 1989 .

[12]  T. J. Mitchell,et al.  Exploratory designs for computational experiments , 1995 .

[13]  Tibor Csendes,et al.  Packing Equal Circles in a Square II. — New Results for up to 100 Circles Using the TAMSASS-PECS Algorithm , 2001 .

[14]  Donald R. Jones,et al.  Efficient Global Optimization of Expensive Black-Box Functions , 1998, J. Glob. Optim..

[15]  Raymond H. Myers,et al.  Response Surface Methodology--Current Status and Future Directions , 1999 .

[16]  N. Oler A Finite Packing Problem , 1961, Canadian Mathematical Bulletin.