On Construction of Sliced Orthogonal Latin Hypercube Designs

Sliced Latin hypercube designs (SLHDs) are considered as appropriate designs while designing computer experiments for collective evaluation of computer models which are related. SLHDs having zero correlation coefficient between any two columns are called sliced orthogonal Latin hypercube designs (SOLHDs). SOLHDs make it possible to independently assess and estimate effects of individual inputs on the response. In this article, we have proposed methods of construction for obtaining SOLHDs with unequal batch sizes. The proposed methods are easy to implement and are capable of accommodating an arbitrary number of factors and flexible in run size.