Application of UQ Principles to Calibration, Sensitivity, and Experimental Design

Wednesday, July 29, 2015 - 1:30pm - 5:00pm
DLR Room 131
Omar Knio (Duke University)
This lecture will outline a probabilistic framework for uncertainty quantification,
and will discuss applications of associated methods and algorithms for the purpose
of calibrating physical parameters from measured data, model selection, and design
of experiments so as to optimize information about specific observables. Concepts
will be illustrated in light of specific applications to nanofluidics and energetic
materials. One of the central goals of the lecture will be highlight the importance
of proper setup of the “UQ problem,” which generally necessitates judicious combination
of information originating from diverse disciplines.