On construction of sliced latin hypercube designs

Sliced Latin hypercube designs (SLHDs) are considered ideal when planning computer experiments for collectively assessing related computer models. SLHDs represent a specific subset of Latin hypercube designs (LHDs), and they can be further subdivided into various slices, each acting as a batch of smaller Latin hypercube designs. In particular, SLHDs with unequal batch sizes find valuable application in designing of computer experiments involving models with varying attributes such as relative cost and accuracy. In this article, we present a simple method for constructing Sliced Latin hypercube designs with unequal batch sizes.