REVIEW 9 cited by
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Physics-Informed Gaussian Process Regression Generalizes Linear PDE Solvers
read the original abstract
Linear partial differential equations (PDEs) are an important, widely applied class of mechanistic models, describing physical processes such as heat transfer, electromagnetism, and wave propagation. In practice, specialized numerical methods based on discretization are used to solve PDEs. They generally use an estimate of the unknown model parameters and, if available, physical measurements for initialization. Such solvers are often embedded into larger scientific models with a downstream application and thus error quantification plays a key role. However, by ignoring parameter and measurement uncertainty, classical PDE solvers may fail to produce consistent estimates of their inherent approximation error. In this work, we approach this problem in a principled fashion by interpreting solving linear PDEs as physics-informed Gaussian process (GP) regression. Our framework is based on a key generalization of the Gaussian process inference theorem to observations made via an arbitrary bounded linear operator. Crucially, this probabilistic viewpoint allows to (1) quantify the inherent discretization error; (2) propagate uncertainty about the model parameters to the solution; and (3) condition on noisy measurements. Demonstrating the strength of this formulation, we prove that it strictly generalizes methods of weighted residuals, a central class of PDE solvers including collocation, finite volume, pseudospectral, and (generalized) Galerkin methods such as finite element and spectral methods. This class can thus be directly equipped with a structured error estimate. In summary, our results enable the seamless integration of mechanistic models as modular building blocks into probabilistic models by blurring the boundaries between numerical analysis and Bayesian inference.
Forward citations
Cited by 9 Pith papers
-
Conditioning Gaussian Processes on Almost Anything
Equivalence between Gaussian processes and linear diffusion models enables general conditioning on arbitrary pointwise likelihoods via ODE dynamics and Monte Carlo guidance approximation.
-
Pressure reconstruction from error-embedded gradient measurements: a Gaussian-process generalization of Green's function integration
Gaussian process regression reconstructs pressure from error-embedded gradients by treating the field as a random process with a fitted correlation kernel, generalizing Green's function integration as its zero-noise l...
-
Structure-Preserving Gaussian Processes Via Discrete Euler-Lagrange Equations
Lagrangian Gaussian Processes use discrete Euler-Lagrange equations to condition GPs, preserving geometric structure for stable dynamics learning from sparse position snapshots without velocities.
-
A unified perspective of Gaussian process approximation for differential equations
A unified Bayesian framework based on derivative matching shows that diverse Gaussian process methods for differential equations are instances of a common probabilistic structure.
-
Error bounds for approximate posteriors from likelihood-informed reduced-order models
Derives error bounds on the root prior-preconditioned Hessian, posterior covariance, and mean for a Petrov-Galerkin reduced-order model, with exact posterior recovery at the intrinsic dimension.
-
Pressure reconstruction from error-embedded gradient measurements: a Gaussian-process generalization of Green's function integration
Gaussian process regression generalizes Green's function integration for pressure reconstruction from error-embedded gradients, matching or exceeding GFI performance on turbulence data while providing calibrated poste...
-
Kernel Learning of PDE Solution Operators
A regularization-based kernel ridge regression approach learns PDE solution operators from physical priors without paired data and provides convergence rates.
-
P-K-GCN: Physics-augmented Koopman-enhanced Graph Convolutional Network for Deep Spatiotemporal Super-resolution
P-K-GCN integrates continuous spline GCN, Koopman linearization, and physics augmentation for spatiotemporal super-resolution on irregular geometries, claiming theoretical error reduction via Rademacher complexity bou...
-
A Dynamical Model for Spatio-Temporal Processes Motivated by Second-Order Partial Differential Equations
A finite-dimensional dynamical model is derived from an infinite-dimensional state-space representation of second-order SPDEs via Galerkin approximation, with space-time covariance and approximation error quantified a...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.