OPTIMAL SECOND DEGREE KRONECKER MODEL MIXTURE EXPERIMENTS WITH TWO, THREE AND FOUR INGREDIENTS
In this study, we present mathematical analysis of an optimal design in the second degree Kronecker model for mixture experiments. We shall purely concentrate on weighted centroid designs in recognition of the completeness result.
We adapt the second degree mixture model postulated by Draper and Pukelsheim [1]. For this model, the full parameter vector is not estimable. We shall select a maximal parameter sub-vector using well constructed parameter matrix.
We use the D- and A- optimality criteria to obtain a set of optimal weighted centroid designs.
optimaldesign,seconddegree Kronecker model, D- and A- optimality, optimal weighted centroid designs.
