CoSy seminar: "Finding optimal and adaptive optimal designs for the continuation-ratio model with nature-inspired metaheuristics"

Everyone is welcome and the first 40 people to register will be treated to a free lunch sandwich. If you do not want lunch, you are still welcome to join.

(Registration opens 7 May)

Abstract: Phase I dose-response (dose-toxicity) studies utilize the  continuation-ratio model to deal with a three-category outcome such as "no-efficacy", "efficacy without toxicity", and "toxicity". This approach allows investigating several objectives such as estimating outcome probabilities as functions of a dose, estimating the most efficacious dose (MED) or a maximum tolerated dose (MTD), etc. The objectives can be transform into a mathematical optimization problem where an optimal solution is found in a form of optimal design, which is represented by a probability measure on a dose interval. In this case, the optimization is performed on a space of probability measures, and finding an optimal design becomes a challenging problem. Also, an optimal design usually depends on model assumptions (parameters) which might be unknown in advance.

In this talk, I’ll give an introduction to the phase I clinical studies, show how the optimal design problem can be solved by a particle swarm optimization algorithm, and demonstrate how adaptive design can be used to approximate the optimal one.      

This is a lecture in the seminar series held by CIM 
(Centre for Interdisciplinary Mathematics).