EcoSta 2017: Start Registration
View Submission - EcoSta2017
A0601
Title: Thoughts on optimal statistical design for environmental risk assessment Authors:  John Stufken - George Mason University (United States) [presenting]
Wanchunzi Yu - Bridgewater State University (United States)
Abstract: A problem that comes up in environmental risk studies is the estimation of a benchmark dose for polluting agents under a dose-response model. A complication is that there is a great deal of model uncertainty. For a study one needs to select doses for the pollutants at which measurements are to be made, which is a design problem. This design problem has not been studied in great detail. We describe the problem and propose a method that helps to identify efficient designs for estimation of the benchmark dose under model uncertainty. The proposed methodology depends on efficient computations, for which a particle swarm optimization (PSO) algorithm can be considered.