Solving and Learning Nonlinear PDEs with Gaussian Processes

DOI10.1016/J.JCP.2021.110668MaRDI QIDQ6363591arXivWikidataFDO

Authors Yifan Chen, Bamdad Hosseini, Houman Owhadi, A. M. Stuart

Publication date 23 March 2021

Published in arXiv

Full work available at URL https://doi.org/10.1016/J.JCP.2021.110668

Mathematics Subject Classification ID

Nonparametric inference (62Gxx) Probabilistic methods, stochastic differential equations (65Cxx) Artificial intelligence (68Txx)

arXiv classification

cs.NAmath.NAstat.ML

Abstract: We introduce a simple, rigorous, and unified framework for solving nonlinear partial differential equations (PDEs), and for solving inverse problems (IPs) involving the identification of parameters in PDEs, using the framework of Gaussian processes. The proposed approach: (1) provides a natural generalization of collocation kernel methods to nonlinear PDEs and IPs; (2) has guaranteed convergence for a very general class of PDEs, and comes equipped with a path to compute error bounds for specific PDE approximations; (3) inherits the state-of-the-art computational complexity of linear solvers for dense kernel matrices. The main idea of our method is to approximate the solution of a given PDE as the maximum a posteriori (MAP) estimator of a Gaussian process conditioned on solving the PDE at a finite number of collocation points. Although this optimization problem is infinite-dimensional, it can be reduced to a finite-dimensional one by introducing additional variables corresponding to the values of the derivatives of the solution at collocation points; this generalizes the representer theorem arising in Gaussian process regression. The reduced optimization problem has the form of a quadratic objective function subject to nonlinear constraints; it is solved with a variant of the Gauss--Newton method. The resulting algorithm (a) can be interpreted as solving successive linearizations of the nonlinear PDE, and (b) in practice is found to converge in a small number of iterations (2 to 10), for a wide range of PDEs. Most traditional approaches to IPs interleave parameter updates with numerical solution of the PDE; our algorithm solves for both parameter and PDE solution simultaneously. Experiments on nonlinear elliptic PDEs, Burgers' equation, a regularized Eikonal equation, and an IP for permeability identification in Darcy flow illustrate the efficacy and scope of our framework.

Cited in

(8)

• Kernel-based Operator Learning: Error Analysis, Budget Allocation, and a Physics-Informed Extension
• On Symmetric Kernel Collocation for Nonlinear PDEs
• A unified perspective of Gaussian process approximation for differential equations
• Cost-Accuracy Trade-offs: Neural Operator vs Classical Numerical Solver
• Gaussian-process surrogate indicators for residual-based adaptive GMsFEM
• Optimal Recovery for Solving Variational Problems
• Improving Learning of Green's Functions Through Kernel Decomposition
• Constrained Probabilistic Numerics: Gaussian Process Solvers for Linear PDEs under Linear Inequality Constraints

This page was built for publication: Solving and Learning Nonlinear PDEs with Gaussian Processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6363591)