We discuss recent mathematical and computational results on uncertainty quantification (UQ) in the presence of uncertainty about the correct probabilistic and physical models. Such UQ problems can be formulated as constrained optimization problems with information acting as the constraints, with consequent optimal assessments of risk, and advantages for interdisciplinary communication and open science. We also report consequences of this point of view for the robustness of Bayesian methods under prior perturbation.
Optimal Uncertainty Quantification: Distributional Robustness Versus Bayesian Brittleness
Manuscript received September 27, 2013; final manuscript received October 18, 2013; published online December 5, 2013. Assoc. Editor: Hengchu Cao.
- Views Icon Views
- Share Icon Share
- Cite Icon Cite
- Search Site
Sullivan, T. J., McKerns, M., Ortiz, M., Owhadi, H., and Scovel, C. (December 5, 2013). "Optimal Uncertainty Quantification: Distributional Robustness Versus Bayesian Brittleness." ASME. J. Med. Devices. December 2013; 7(4): 040920. https://doi.org/10.1115/1.4025786
Download citation file:
- Ris (Zotero)
- Reference Manager