Bayesian numerical methods for nonlinear partial differential equations
From MaRDI portal
Publication:2058795
Abstract: The numerical solution of differential equations can be formulated as an inference problem to which formal statistical approaches can be applied. However, nonlinear partial differential equations (PDEs) pose substantial challenges from an inferential perspective, most notably the absence of explicit conditioning formula. This paper extends earlier work on linear PDEs to a general class of initial value problems specified by nonlinear PDEs, motivated by problems for which evaluations of the right-hand-side, initial conditions, or boundary conditions of the PDE have a high computational cost. The proposed method can be viewed as exact Bayesian inference under an approximate likelihood, which is based on discretisation of the nonlinear differential operator. Proof-of-concept experimental results demonstrate that meaningful probabilistic uncertainty quantification for the unknown solution of the PDE can be performed, while controlling the number of times the right-hand-side, initial and boundary conditions are evaluated. A suitable prior model for the solution of the PDE is identified using novel theoretical analysis of the sample path properties of Mat'{e}rn processes, which may be of independent interest.
Recommendations
- Numerical Gaussian processes for time-dependent and nonlinear partial differential equations
- A role for symmetry in the Bayesian solution of differential equations
- Probabilistic solutions to ordinary differential equations as nonlinear Bayesian filtering: a new perspective
- Bayesian Probabilistic Numerical Methods
- Variational Bayes' Method for Functions with Applications to Some Inverse Problems
Cites work
- A Method for the Spatial Discretization of Parabolic Equations in One Space Variable
- A modern retrospective on probabilistic numerics
- A probabilistic model for the numerical solution of initial value problems
- A role for symmetry in the Bayesian solution of differential equations
- Adaptive step-size selection for state-space probabilistic differential equation solvers
- Bayesian calibration of computer models. (With discussion)
- Bayesian numerical homogenization
- Bayesian ODE solvers: the maximum a posteriori estimate
- Bayesian Probabilistic Numerical Methods
- Bayesian probabilistic numerical methods in time-dependent state estimation for industrial hydrocyclone equipment
- Bayesian solution uncertainty quantification for differential equations
- Compactly supported correlation functions
- Compression, inversion, and approximate PCA of dense kernel matrices at near-linear computational complexity
- Convergence rates of Gaussian ODE filters
- Estimating shape constrained functions using Gaussian processes
- Finite Volume Methods for Hyperbolic Problems
- Gaussian measure in Hilbert space and applications in numerical analysis
- Gaussian processes for machine learning.
- scientific article; zbMATH DE number 1001278 (Why is no real title available?)
- scientific article; zbMATH DE number 1804115 (Why is no real title available?)
- scientific article; zbMATH DE number 4098703 (Why is no real title available?)
- scientific article; zbMATH DE number 88931 (Why is no real title available?)
- scientific article; zbMATH DE number 3591122 (Why is no real title available?)
- scientific article; zbMATH DE number 1181255 (Why is no real title available?)
- scientific article; zbMATH DE number 2152342 (Why is no real title available?)
- scientific article; zbMATH DE number 802686 (Why is no real title available?)
- scientific article; zbMATH DE number 3045283 (Why is no real title available?)
- scientific article; zbMATH DE number 3078316 (Why is no real title available?)
- Integration of covariance kernels and stationarity
- Interpolation of spatial data. Some theory for kriging
- Mathematical methods for financial markets.
- Maximum likelihood estimation and uncertainty quantification for Gaussian process approximation of deterministic functions
- Mechanistic hierarchical Gaussian processes
- Numerical Gaussian processes for time-dependent and nonlinear partial differential equations
- Operator splitting
- Probabilistic numerics and uncertainty in computations
- Probabilistic solutions to ordinary differential equations as nonlinear Bayesian filtering: a new perspective
- Regularity of the sample paths of a general second order random field
- Sample properties of random fields. III: Differentiability
- Solving differential equations with radial basis functions: Multilevel methods and smoothing
- Solving PDEs with radial basis functions
- The sizes of compact subsets of Hilbert space and continuity of Gaussian processes
Cited in
(24)- Inferring solutions of differential equations using noisy multi-fidelity data
- Statistical analysis of differential equations: introducing probability measures on numerical solutions
- Gaussian functional regression for linear partial differential equations
- Randomised one-step time integration methods for deterministic operator differential equations
- Approximate Bayesian model inversion for PDEs with heterogeneous and state-dependent coefficients
- Probabilistic solutions to ordinary differential equations as nonlinear Bayesian filtering: a new perspective
- A new network approach to Bayesian inference in partial differential equations
- Numerical Gaussian processes for time-dependent and nonlinear partial differential equations
- Bayesian approach to a nonlinear inverse problem for a time-space fractional diffusion equation
- Error control in the numerical posterior distribution in the Bayesian UQ analysis of a semilinear evolution PDE
- Statistical finite elements for misspecified models
- On Bayesian data assimilation for PDEs with ill-posed forward problems
- Bayesian solution uncertainty quantification for differential equations
- Comment on article by Chkrebtii, Campbell, Calderhead and Girolami
- Bayesian Probabilistic Numerical Methods
- Comment on article by Chkrebtii, Campbell, Calderhead and Girolami
- Machine learning for a class of partial differential equations with multi-delays based on numerical Gaussian processes
- Sparse Gaussian processes for solving nonlinear PDEs
- A Bayesian approach to modeling finite element discretization error
- Parameter inference based on Gaussian processes informed by nonlinear partial differential equations
- Error analysis of kernel/GP methods for nonlinear and parametric PDEs
- Convergence rates of non-stationary and deep Gaussian process regression
- Dynamic Bayesian learning for spatiotemporal mechanistic models
- Gaussian measures conditioned on nonlinear observations: consistency, MAP estimators, and simulation
This page was built for publication: Bayesian numerical methods for nonlinear partial differential equations
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2058795)
