Designs for Large-Scale Simulation Experiments, with Applications to Defense and Homeland Security

Abstract

Computer experimentation is integral to modern scientific research, national defense, industry and manufacturing, and in public policy debates. Traditional design of experiments (DOE) focuses on small-scale experimentation, whether dealing with experiments involving physical systems or computer models. In contrast, computer models tend to be extremely complex, often with thousands of factors and many sources of uncertainty. Historically, even if experimental designs have been used, they have typically been applied to only a handful of factors—even for computer models having hundreds or thousands of inputs. This suggests that more modelers and analysts need to be aware of the power of experimental design—especially the recent breakthroughs in large-scale experimental designs that enable us to understand the impact of many factors and their intricate interactions on model outcomes.