Weighted Simplex Centroid Mixture Experiments for third order Becker’s models: The R-optimal approach

Abstract

The R-optimality criterion, suggested as a substitute for the widely used D-optimality criteria, is employed in experimental designs when the primary goal is to create a rectangular confidence region. This study explores R-optimal designs concerning third order Becker’s models and calculates weights for values of $3\le q\le 10$. Additionally, it compares and contrasts the D-optimal and R-optimal designs.