Over the last 20 years, the widespread availability of optimal design algorithms in statistical computing packages has revolutionized the application of design of experiments (DOE). More recently, the emphasis has shifted from very general optimization methods which are computationally manageable, to the development of optimal designs which are tailored to the specific research questions experiments are intended to answer. There is now a push towards unifying algorithmic approaches to design construction, with the more traditional combinatorial approaches, to ensure that designs can truly be optimized for the relevant applications.
Areas covered in this theme will include discrete choice experiments, response surface experiments with hard-to-set factors, nonlinear design in biological kinetics and design of measurement systems for communication networks.
