Applied Math Graduate Student Seminar:John Darges, NC State, Goal-Oriented Variance-Based Sensitivity Analysis for Uncovering Prior Hyperparameter Importance in Bayesian Inverse Problems s
SAS 4201The formulation of Bayesian inverse problems involves choosing prior distributions; choices that seem equally reasonable may lead to significantly different conclusions. We develop a computational approach to better understand the impact of the hyperparameters defining the prior on the posterior statistics of the quantities of interest. Our approach relies on global sensitivity analysis (GSA) of…