We propose a model validation procedure that consists of a prediction error identi6cation experiment with a full order model. It delivers a parametric uncertainty ellipsoid and a corresponding set of parameterized transfer functions, which we call prediction error (PE) unceminty set. Such uncertainty set differs from the classical uncertainty descriptions used in robust control analysis and des...

We study statistical inference and distributionally robust solution methods for stochastic optimization problems, focusing on confidence intervals optimal values solutions that achieve exact coverage asymptotically. develop a generalized empirical likelihood framework—based distributional uncertainty sets constructed from nonparametric f-divergence balls—for Hadamard differentiable functionals,...