Abstract:
|
The tremendous gains in computing power and the advances in algorithms, together with the availability of data, provide scientists with unprecedented capability for quantitative analysis. But the mathematical and statistical models used to construct and interpret this quantitative data rarely account for every feature that might influence the outcome. It is, therefore, of great importance to understand the uncertainties inherent in the models. In statistics this process is known as Model Uncertainty; in engineering it is called Uncertainty Quantification.
During 2018-2019, the Statistical and Applied Mathematical Sciences Institute organized a year-long program on Model Uncertainty: Mathematical and Statistical. This talk will provide an overview of that program and the collaborations it fostered, and will briefly summarize some of research outputs from the participants. We will discuss the operations of one of the research working groups, and illustrate some of the novel ideas generated from that group.
|