Latin hypercube designs with controlled correlations and multi - dimensional stratification
نویسندگان
چکیده
Various methods have been proposed to construct Latin hypercube designs with small correlations. Orthogonal arrays have been used to construct Latin hypercube designs with multidimensional stratification. To integrate these two ideas, we propose a method to construct Latin hypercube designs with both controlled correlations and multi-dimensional stratification. For numerical integration, the constructed designs not only filter out lower-dimensional variance components as effectively as ordinary orthogonal array-based Latin hypercube designs, but also filter out bilinear terms more effectively. The proposed construction method entails no iterative searches. Sampling properties of the constructed designs are derived. Examples are given to illustrate the proposed construction method and the theoretical results.
منابع مشابه
Sliced Latin hypercube designs via orthogonal arrays
Computer experiments are becoming increasingly popular in studying complex real world systems. A special class of sliced Latin hypercube design is proposed in this paper. Such designs are particularly suitable for computer experiments with both qualitative and quantitative factors, multi-fidelity computer experiments, cross-validation and data pooling. The resulting sliced Latin hypercube desig...
متن کاملOrthogonal-maximin Latin Hypercube Designs
A randomly generated Latin hypercube design (LHD) can be quite structured: the variables may be highly correlated or the design may not have good space-filling properties. There are procedures for finding good LHDs by minimizing the pairwise correlations or by maximizing the inter-site distances. In this article we show that these two criteria need not be in close agreement. We propose a multi-...
متن کاملFast Generation of Space-filling Latin Hypercube Sample Designs
Latin Hypercube Sampling (LHS) and Jittered Sampling (JS) both achieve better convergence than standard Monte Carlo Sampling (MCS) by using stratification to obtain a more uniform selection of samples, although LHS and JS use different stratification strategies. The “Koksma-Hlawka-like inequality” bounds the error in a computed mean in terms of the sample design’s discrepancy, which is a common...
متن کاملTwo-dimensional minimax Latin hypercube designs
We investigate minimax Latin hypercube designs in two dimensions for several distance measures. For the `-distance we are able to construct minimax Latin hypercube designs of n points, and to determine the minimal covering radius, for all n. For the `1-distance we have a lower bound for the covering radius, and a construction of minimax Latin hypercube designs for (infinitely) many values of n....
متن کاملSpace-filling Latin hypercube designs for computer experiments
In the area of computer simulation Latin hypercube designs play an important role. In this paper the class of maximin Latin hypercube designs is considered. Up to now only several two-dimensional designs and designs for some small number of points are known for this class. Using periodic designs and simulated annealing we extend the known results and construct approximate maximin Latin hypercub...
متن کامل