Approximate Bayesian model inversion for PDEs with heterogeneous and state-dependent coefficients
From MaRDI portal
Publication:2222341
DOI10.1016/J.JCP.2019.06.010zbMATH Open1453.62410arXiv1902.06718OpenAlexW2911737975WikidataQ114163517 ScholiaQ114163517MaRDI QIDQ2222341FDOQ2222341
Authors: David A. Barajas-Solano, Alexandre M. Tartakovsky 
Publication date: 26 January 2021
Published in: Journal of Computational Physics (Search for Journal in Brave)
Abstract: We present two approximate Bayesian inference methods for parameter estimation in partial differential equation (PDE) models with space-dependent and state-dependent parameters. We demonstrate that these methods provide accurate and cost-effective alternatives to Markov Chain Monte Carlo simulation. We assume a parameterized Gaussian prior on the unknown functions, and approximate the posterior density by a parameterized multivariate Gaussian density. The parameters of the prior and posterior are estimated from sparse observations of the PDE model's states and the unknown functions themselves by maximizing the evidence lower bound (ELBO), a lower bound on the log marginal likelihood of the observations. The first method, Laplace-EM, employs the expectation maximization algorithm to maximize the ELBO, with a Laplace approximation of the posterior on the E-step, and minimization of a Kullback-Leibler divergence on the M-step. The second method, DSVI-EB, employs the doubly stochastic variational inference (DSVI) algorithm, in which the ELBO is maximized via gradient-based stochastic optimization, with nosiy gradients computed via simple Monte Carlo sampling and Gaussian backpropagation. We apply these methods to identifying diffusion coefficients in linear and nonlinear diffusion equations, and we find that both methods provide accurate estimates of posterior densities and the hyperparameters of Gaussian priors. While the Laplace-EM method is more accurate, it requires computing Hessians of the physics model. The DSVI-EB method is found to be less accurate but only requires gradients of the physics model.
Full work available at URL: https://arxiv.org/abs/1902.06718
Recommendations
- Variational Bayes' Method for Functions with Applications to Some Inverse Problems
- Variational Bayesian approximation of inverse problems using sparse precision matrices
- A hierarchical Bayesian setting for an inverse problem in linear parabolic PDEs with noisy boundary conditions
- Bayesian numerical methods for nonlinear partial differential equations
- Stein variational reduced basis Bayesian inversion
Bayesian inference (62F15) Dependence of solutions to PDEs on initial and/or boundary data and/or on parameters of PDEs (35B30)
Cites Work
- Automatic differentiation variational inference
- The no-U-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo
- Pattern recognition and machine learning.
- Gaussian Kullback-Leibler approximate inference
- Ensemble samplers with affine invariance
- Inverse problems: a Bayesian perspective
- Data Assimilation
- A regularizing Levenberg - Marquardt scheme, with applications to inverse groundwater filtration problems
- Simple statistical gradient-following algorithms for connectionist reinforcement learning
- Title not available (Why is that?)
- Bayesian and variational Bayesian approaches for flows in heterogeneous random media
- Hierarchical Bayesian inference for ill-posed problems via variational method
- Improved Laplace approximation for marginal likelihoods
- Machine learning of linear differential equations using Gaussian processes
- Sparse variational Bayesian approximations for nonlinear inverse problems: applications in nonlinear elastography
- Numerical Gaussian processes for time-dependent and nonlinear partial differential equations
- A variational Bayesian approach for inverse problems with skew-\(t\) error distributions
Cited In (15)
- VI-DGP: a variational inference method with deep generative prior for solving high-dimensional inverse problems
- Multifidelity multilevel Monte Carlo to accelerate approximate Bayesian parameter inference for partially observed stochastic processes
- Reduced-order model-based variational inference with normalizing flows for Bayesian elliptic inverse problems
- Proper Generalized Decomposition model reduction in the Bayesian framework for solving inverse heat transfer problems
- Physics-informed Karhunen-Loéve and neural network approximations for solving inverse differential equation problems
- Variational inference at glacier scale
- On-the-fly construction of surrogate constitutive models for concurrent multiscale mechanical analysis through probabilistic machine learning
- Data-driven learning of differential equations: combining data and model uncertainty
- Learning and meta-learning of stochastic advection–diffusion–reaction systems from sparse measurements
- A new network approach to Bayesian inference in partial differential equations
- A hierarchical Bayesian setting for an inverse problem in linear parabolic PDEs with noisy boundary conditions
- Variational Bayesian approximation of inverse problems using sparse precision matrices
- Physics-informed machine learning with conditional Karhunen-Loève expansions
- On Bayesian data assimilation for PDEs with ill-posed forward problems
- Approximation of Bayesian Inverse Problems for PDEs
Uses Software
This page was built for publication: Approximate Bayesian model inversion for PDEs with heterogeneous and state-dependent coefficients
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2222341)
