Answers are grounded in this paper and cite their sources.
This paper discusses an algorithm for constructing mixed-level screening designs (MLSDs) by augmenting some columns of a definitive screening designs (DSDs) with additional two-level columns. The constructed designs have the quadratic effects being orthogonal to main effects. The algorithm is used to construct designs with the number of runs being equal to p (i.e., the number of model parameters) for even p or p + 1 for odd p. The performance of these small DSD-based MLSDs in terms of the D-efficiency is evaluated against the 60 small MLSDs of Yang et al. (2014).
No takes yet. Share an insight, caveat, or question.
Nguyen et al. (2016) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: