Jian-Feng Yang

Construction of sliced (nearly) orthogonal Latin hypercube designs

Abstract Sliced Latin hypercube designs are very useful for running a computer model in batches, ensembles of multiple computer models, computer experiments with qualitative and quantitative factors, cross-validation and data pooling. However, the presence of highly correlated columns makes the data analysis intractable. In this paper, a construction method for sliced (nearly) orthogonal Latin hypercube designs is developed. The resulting designs have flexible sizes and most are new. With the orthogonality or near orthogonality being guaranteed, the space-filling property of the resulting designs is also improved. Examples are provided for illustrating the proposed method.

Computer Experiments With Both Qualitative and Quantitative Variables

Computer experiments have received a great deal of attention in many fields of science and technology. Most literature assumes that all the input variables are quantitative. However, researchers often encounter computer experiments involving both qualitative and quantitative variables (BQQV). In this article, a new interface on design and analysis for computer experiments with BQQV is proposed. The new designs are one kind of sliced Latin hypercube designs with points clustered in the design region and possess good uniformity for each slice. For computer experiments with BQQV, such designs help to measure the similarities among responses of different level-combinations in the qualitative variables. An adaptive analysis strategy intended for the proposed designs is developed. The proposed strategy allows us to automatically extract information from useful auxiliary responses to increase the precision of prediction for the target response. The interface between the proposed design and the analysis strategy is demonstrated to be effective via simulation and a real-life example from the food engineering literature. Supplementary materials for this article are available online.

Construction of Optimal Blocking Schemes for Robust Parameter Designs

Robust parameter design (RPD) is an important issue in experimental designs. If all experimental runs cannot be performed under homogeneous conditions, blocking the units is effective. In this paper, we obtain the correspondence relation between fractional factorial RPDs and the blocking schemes for full factorial RPDs. In addition, we provide a construction of optimal blocking schemes that make all main effects and control-by-noise two-factor interactions estimable.

Some Results on Fractional Factorial Split-Plot Designs with Multi-Level Factors

Fractional factorial split-plot (FFSP) designs have received much attention in recent years. In this article, the matrix representation for FFSP designs with multi-level factors is first developed, which is an extension of the one proposed by Bingham and Sitter (1999b) for the two-level case. Based on this representation, periodicity results of maximum resolution and minimum aberration for such designs are derived. Differences between FFSP designs with multi-level factors and those with two-level factors are highlighted.

Construction of nearly uniform designs on irregular regions

ABSTRACT The uniform design is a kind of important experimental design which has great practical value in production and living. Most existing literatures on this topic focus on the construction of uniform designs on regular regions. However, because of the complexity of practical situations, the irregular design region is more common in real life. In this paper, an algorithm is proposed to the construction of nearly uniform designs on irregular regions. The basic idea is to make use of uniform designs on a larger regular region with the irregular region being a subregion. Some theoretical justifications on the proposed algorithm are provided. Both the comparisons with the existing results and a real-life example show that our proposed algorithm is effective.

Construction of regular 2n41 designs with general minimum lower-order confounding

ABSTRACT Mixed-level designs, especially two- and four-level designs, are very useful in practice. In the last two decades, there are quite a few literatures investigating the selection of this kind of optimal designs. Recently, the general minimum lower-order confounding (GMC) criterion (Zhang et al., 2008) gave a new approach for choosing optimal factorials. It is proved that the GMC designs are more powerful than other criteria in the widely practical situations. In this paper, we extend the GMC theory to the mixed-level designs. Under the theory we establish a new criterion for choosing optimal regular two- and four-level designs. Further, a construction method is proposed to obtain all the 2n41 GMC designs with N/4 + 1 ⩽ n + 2 ⩽ 5N/16, where N is the number of runs and n is the number of two-level factors.

Construction of some mixed two- and four-level regular designs with GMC criterion

ABSTRACT General minimum lower-order confounding (GMC) criterion is to choose optimal designs, which are based on the aliased effect-number pattern (AENP). The AENP and GMC criterion have been developed to form GMC theory. Zhang et al. (2015) introduced GMC 2n4m criterion for choosing optimal designs and constructed all GMC 2n41 designs with N/4 + 1 ⩽ n + 2 ⩽ 5N/16. In this article, we analyze the properties of 2n41 designs and construct GMC 2n41 designs with 5N/16 + 1 ⩽ n + 2 < N − 1, where n and N are, respectively, the numbers of two-level factors and runs. Further, GMC 2n41 designs with 16-run, 32-run are tabulated.