Low-rank tensor reconstruction of concentrated densities with application to Bayesian inversion
From MaRDI portal
Publication:2128063
Bayesian inversionlow-rankpartial differential equations with random coefficientstensor trainuncertainty quantificationVMC
Computational methods for stochastic equations (aspects of stochastic analysis) (60H35) Computational methods for problems pertaining to statistics (62-08) Bayesian inference (62F15) Density estimation (62G07) Spectral, collocation and related methods for boundary value problems involving PDEs (65N35)
Abstract: Transport maps have become a popular mechanic to express complicated probability densities using sample propagation through an optimized push-forward. Beside their broad applicability and well-known success, transport maps suffer from several drawbacks such as numerical inaccuracies induced by the optimization process and the fact that sampling schemes have to be employed when quantities of interest, e.g. moments are to compute. This paper presents a novel method for the accurate functional approximation of probability density functions (PDF) that copes with those issues. By interpreting the pull-back result of a target PDF through an inexact transport map as a perturbed reference density, a subsequent functional representation in a more accessible format allows for efficient and more accurate computation of the desired quantities. We introduce a layer-based approximation of the perturbed reference density in an appropriate coordinate system to split the high-dimensional representation problem into a set of independent approximations for which separately chosen orthonormal basis functions are available. This effectively motivates the notion of h- and p-refinement (i.e. ``mesh size and polynomial degree) for the approximation of high-dimensional PDFs. To circumvent the curse of dimensionality and enable sampling-free access to certain quantities of interest, a low-rank reconstruction in the tensor train format is employed via the Variational Monte Carlo method. An a priori convergence analysis of the developed approach is derived in terms of Hellinger distance and the Kullback-Leibler divergence. Applications comprising Bayesian inverse problems and several degrees of concentrated densities illuminate the (superior) convergence in comparison to Monte Carlo and Markov-Chain Monte Carlo methods.
Recommendations
- Sampling-free Bayesian inversion with adaptive hierarchical tensor representations
- Approximation and sampling of multivariate probability distributions in the tensor train decomposition
- Rank bounds for approximating Gaussian densities in the tensor-train format
- Deep composition of tensor-trains using squared inverse Rosenblatt transports
- Low-rank independence samplers in hierarchical Bayesian inverse problems
Cites work
- N-term Wiener chaos approximation rates for elliptic PDEs with lognormal Gaussian random inputs
- A continuous analogue of the tensor-train decomposition
- A literature survey of low-rank tensor approximation techniques
- A multiscale strategy for Bayesian inference using transport maps
- A Sparse Grid Stochastic Collocation Method for Partial Differential Equations with Random Input Data
- A Stochastic Collocation Method for Elliptic Partial Differential Equations with Random Input Data
- A Stochastic Collocation Method for Elliptic Partial Differential Equations with Random Input Data
- Adaptive construction of surrogates for the Bayesian solution of inverse problems
- Adaptive stochastic Galerkin FEM
- Adaptive stochastic Galerkin FEM for lognormal coefficients in hierarchical tensor representations
- Adaptive stochastic Galerkin FEM with hierarchical tensor representations
- Approximation and sampling of multivariate probability distributions in the tensor train decomposition
- Approximation rates for the hierarchical tensor format in periodic Sobolev spaces
- Bayesian inference with optimal maps
- Black box approximation of tensors in hierarchical Tucker format
- Black box low tensor-rank approximation using fiber-crosses
- scientific article; zbMATH DE number 2101400 (Why is no real title available?)
- scientific article; zbMATH DE number 840151 (Why is no real title available?)
- Inverse problems: a Bayesian perspective
- Multi-element probabilistic collocation method in high dimensions
- Multilevel Markov Chain Monte Carlo
- Non-intrusive tensor reconstruction for high-dimensional random PDEs
- Normalizing flows for probabilistic modeling and inference
- On a Metropolis-Hastings importance sampling estimator
- On the construction of sparse tensor product spaces
- On the convergence of the Laplace approximation and noise-level-robustness of Laplace-based Monte Carlo methods for Bayesian inverse problems
- On the Convergence Rate of Generalized Fourier Expansions
- On the Lebesgue constant of subperiodic trigonometric interpolation
- Optimal Transport
- Optimal transport for applied mathematicians. Calculus of variations, PDEs, and modeling
- Optimal weighted least-squares methods
- Orthogonal polynomials of several variables
- Parametric PDEs: sparse or low-rank approximations?
- Sampling-free Bayesian inversion with adaptive hierarchical tensor representations
- Scaling limits in computational Bayesian inversion
- Sparse tensor discretizations of high-dimensional parametric and stochastic PDEs
- Sparse-grid, reduced-basis Bayesian inversion: nonaffine-parametric nonlinear equations
- Statistical and computational inverse problems.
- Tensor networks and hierarchical tensors for the solution of high-dimensional partial differential equations
- Tensor spaces and numerical tensor calculus
- Tensor-train decomposition
- Transport map accelerated Markov chain Monte Carlo
- TT-cross approximation for multidimensional arrays
- Variational Monte Carlo -- bridging concepts of machine learning and high-dimensional partial differential equations
Cited in
(12)- Deep composition of tensor-trains using squared inverse Rosenblatt transports
- Approximation and sampling of multivariate probability distributions in the tensor train decomposition
- Sampling-free Bayesian inversion with adaptive hierarchical tensor representations
- Rank bounds for approximating Gaussian densities in the tensor-train format
- Adaptive Nonintrusive Reconstruction of Solutions to High-Dimensional Parametric PDEs
- An hp‐adaptive multi‐element stochastic collocation method for surrogate modeling with information re‐use
- Scalable conditional deep inverse Rosenblatt transports using tensor trains and gradient-based dimension reduction
- Deep Importance Sampling Using Tensor Trains with Application to a Priori and a Posteriori Rare Events
- Approximation in the extended functional tensor train format
- Reverse diffusion sampling with tensor train approximations of Hamilton-Jacobi-Bellman equations
- The tensor-train stochastic finite volume method for uncertainty quantification
- Tensor-Based Reduced-Order Modeling for Optimization-Based Inverse Problems
This page was built for publication: Low-rank tensor reconstruction of concentrated densities with application to Bayesian inversion
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2128063)
