LOCATION: 214 Atanasoff
SPEAKER: Dorin Drignei, Department of Statistics, Iowa State University
TITLE: Statistical Analysis of Multivariate Computer Output
ABSTRACT:
| Many scientific investigations rely on computer models for simulating plausible real situations. This is especially useful when physical experimentation is too expensive or even impossible. In trying to describe the complexities of reality, some computer models are themselves very complex and are therefore expensive to run, in terms of computational resources and time. In response to some of these issues, a recent approach proposes to use statistical models as less computationally demanding surrogates of such complex computer models. More precisely, based on a number of selected runs of the computer model, one builds a statistical model to predict the output of the computer model for untried runs. These statistical surrogates do not exactly match the computer model output in a new situation, but they have the capability to describe the associated uncertainty. Ideally, the completed statistical model would not require as many computational resources as the original computer model. In this talk we propose a statistical model for multivariate computer output of finite difference solvers of differential equations. This statistical model makes use of the underlying code information and, as a result, is second-order non-stationary. The Lotka-Volterra competing species differential system will be used as an example to illustrate the methods. It will be shown that the statistical model proposed here is more accurate than a statistical model that extends directly the existing scalar methodology to the multivariate case.
|