Using numerical optimization techniques for sampling in statistical inverse problems

Authors

John M. Bardsley, Department of Mathematical Sciences

Document Type

Presentation Abstract

Presentation Date

3-2-2015

Abstract

Many solution methods for inverse problems compute the maximum a posteriori (MAP) estimator, or equivalently, the regularized solution, by solving an optimization problem. Uncertainty quantification (UQ), on the other hand, typically requires sampling from the Bayesian posterior density function. In this talk, we bring these two ideas together and present posterior sampling methods that make use of existing algorithms for computing regularized solutions/MAP estimators. Theoretically correct samplers for both linear and nonlinear inverse problems will be presented.

Additional Details

Monday, March 2, 2015 at 3:10 p.m. in Math 103
4:00 p.m. Refreshments in Math Lounge 109

Recommended Citation

Bardsley, John M., "Using numerical optimization techniques for sampling in statistical inverse problems" (2015). Colloquia of the Department of Mathematical Sciences. 479.
https://scholarworks.umt.edu/mathcolloquia/479