Analysis of a Computational Framework for Bayesian Inverse Problems: Ensemble Kalman Updates and MAP Estimators under Mesh Refinement
From MaRDI portal
Publication:6131418
DOI10.1137/23M1567035arXiv2304.09933OpenAlexW4391476783MaRDI QIDQ6131418FDOQ6131418
Authors: Daniel Sanz-Alonso 
Publication date: 5 April 2024
Published in: SIAM/ASA Journal on Uncertainty Quantification (Search for Journal in Brave)
Abstract: This paper analyzes a popular computational framework to solve infinite-dimensional Bayesian inverse problems, discretizing the prior and the forward model in a finite-dimensional weighted inner product space. We demonstrate the benefit of working on a weighted space by establishing operator-norm bounds for finite element and graph-based discretizations of Mat'ern-type priors and deconvolution forward models. For linear-Gaussian inverse problems, we develop a general theory to characterize the error in the approximation to the posterior. We also embed the computational framework into ensemble Kalman methods and MAP estimators for nonlinear inverse problems. Our operator-norm bounds for prior discretizations guarantee the scalability and accuracy of these algorithms under mesh refinement.
Full work available at URL: https://arxiv.org/abs/2304.09933
Recommendations
- A computational framework for infinite-dimensional Bayesian inverse problems. I: The linearized case, with application to global seismic inversion
- Data-driven forward discretizations for Bayesian inversion
- Ensemble Kalman filter for multiscale inverse problems
- Hierarchical ensemble Kalman methods with sparsity-promoting generalized gamma hyperpriors
- Well-posed Bayesian inverse problems and heavy-tailed stable quasi-Banach space priors
Bayesian inference (62F15) Analysis of algorithms and problem complexity (68Q25) Numerical methods for inverse problems for initial value and initial-boundary value problems involving PDEs (65M32) PDEs in connection with statistics (35Q62)
Cites Work
- Interpolation of spatial data. Some theory for kriging
- The Rational SPDE Approach for Gaussian Random Fields With General Smoothness
- Gaussian processes for machine learning.
- Statistical and computational inverse problems.
- Whittle-Matérn priors for Bayesian statistical inversion with applications in electrical impedance tomography
- An introduction to matrix concentration inequalities
- Importance sampling: intrinsic dimension and computational cost
- An introduction to \(\Gamma\)-convergence
- Infinite-dimensional dynamical systems in mechanics and physics.
- Infinite-dimensional dynamical systems. An introduction to dissipative parabolic PDEs and the theory of global attractors
- Inverse problems: a Bayesian perspective
- Title not available (Why is that?)
- The Mathematical Theory of Finite Element Methods
- Title not available (Why is that?)
- A handbook of \(\Gamma\)-convergence
- Well-posed stochastic extensions of ill-posed linear problems
- Inverse Problem Theory and Methods for Model Parameter Estimation
- Semi-supervised learning on Riemannian manifolds
- Title not available (Why is that?)
- MCMC methods for functions: modifying old algorithms to make them faster
- A stochastic collocation approach to Bayesian inference in inverse problems
- Approximation of Bayesian Inverse Problems for PDEs
- A computational framework for infinite-dimensional Bayesian inverse problems. I: The linearized case, with application to global seismic inversion
- Ensemble Kalman methods for inverse problems
- MAP estimators and their consistency in Bayesian nonparametric inverse problems
- A function space HMC algorithm with second order Langevin diffusion limit
- A graph discretization of the Laplace-Beltrami operator
- Learning Theory
- A computational framework for infinite-dimensional Bayesian inverse problems. II: stochastic Newton MCMC with application to ice sheet flow inverse problems
- Bayesian inverse problems for functions and applications to fluid mechanics
- On the well-posedness of Bayesian inverse problems
- Hypermodels in the Bayesian imaging framework
- Iterative ensemble Kalman methods: a unified perspective with some new variants
- Iterative regularization for ensemble data assimilation in reservoir models
- Maximum a posteriori probability estimates in infinite-dimensional Bayesian inverse problems
- Error estimates for spectral convergence of the graph Laplacian on random geometric graphs toward the Laplace-Beltrami operator
- The SPDE approach to Matérn fields: graph representations
- Convergence of the square root ensemble Kalman filter in the large ensemble limit
- Functional analysis: spectral theory
- On the local Lipschitz stability of Bayesian inverse problems
- Continuum limits of posteriors in graph Bayesian inverse problems
- Kernel Methods for Bayesian Elliptic Inverse Problems on Manifolds
- Calibrate, emulate, sample
- On the consistency of graph-based Bayesian semi-supervised learning and the scalability of sampling algorithms
- Data-driven forward discretizations for Bayesian inversion
- Inverse Problems and Data Assimilation
- Graph-based prior and forward models for inverse problems on manifolds with boundaries
- Bayesian Inverse Problems Are Usually Well-Posed
- Maximum a posteriori estimators in ℓp are well-defined for diagonal Gaussian priors
This page was built for publication: Analysis of a Computational Framework for Bayesian Inverse Problems: Ensemble Kalman Updates and MAP Estimators under Mesh Refinement
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6131418)
