Using Animal Instincts to Design Efficient Experiments for Biomedical Studies
Weng Kee Wong (UCLA)
Friday 3rd July, 2015 15:00-16:00 Maths 203
Nature-inspired meta-heuristic algorithms are increasingly studied and used in computer science and engineering disciplines to solve high-dimensional complex optimization problems. These general optimization methods are simple to implement, very flexible and able to find an optimal or a nearly optimal solution quickly. In addition, they usually do not require any assumption on the function to be optimized and are generally not impeded by the technical or physical constraints of the problem.
I will first give a brief overview of optimal design methodology and recent advances in the field before I demonstrate the usefulness of some of these algorithms for generating different types of optimal designs for dose response studies. Specific applications include constructing efficient designs for estimating interesting features in compartmental models, Emax models and continuation ratio models with bivariate binary responses. If time permits, I will also discuss search algorithms for finding maximin or minimax types of optimal designs for a nonlinear mod