Monday, June 18, 20121:00 PM - 2:00 PMCNLS Conference Room (TA-3, Bldg 1690)|
Separating intrinsic and parametric uncertainty for UQ in stochastic epidemic models
Kyle HickmanTulane University
The first half of this talk will serve as an introduction to the field of uncertainty quantification. We will discuss motivation for the recent development of this field and some of the main problems of interest. As an example we will look at work in progress to quantify the uncertainty in stochastic models of epidemics. Difficulties that arise in UQ for these types of models will be shown and techniques to overcome these difficulties will be explained. This will involve an overview of Karhunen-Loeve and polynomial chaos decompositions as well as Gaussian process regressions. Advantages and shortcomings of these methods will be highlighted and open problems in the field will be shown.
Host: Humberto C Godinez Vazquez, Mathematical Modeling and Analysis Theoretical Division, 5-9188