Summary
Solving equations that describe how heat spreads in space and time is tricky when the heat source has random noise. The authors present a new way to learn these heat patterns from limited and noisy data by combining mathematical equations with sample data intelligently. Their method uses energy concepts from physics to guide learning, leading to better predictions and understanding of how space and time affect accuracy differently. They also provide mathematical guarantees on how well their approach can estimate solutions and generalize to unseen data. Tests show their method outperforms some existing neural network techniques in predicting heat and related physical systems.
What this means in practice
- •For engineering simulation teams: Improve prediction accuracy for heat diffusion models using fewer noisy measurements by applying physics-based energy regularization.
- •For numerical methods developers: Design algorithms that incorporate spatial and temporal sampling strategies optimized for parabolic PDE parameter estimation under noise.
Abstract
Data and the governing partial differential equations (PDEs) of continuous-time dynamical systems provide rich structure for statistical learning. A central challenge is how to integrate data with PDE constraints in a principled way to improve generalization. In this paper, we characterize the statistical estimation and generalization properties of learning solutions to the heat equation with a noisy source function from finite, randomly sampled data. We first derive a novel variational formulation, which we characterize as the minimizer of loss function for an energy regularized and ellipticity debiased Sobolev regression. By formulating the learning task as a supervised problem over space-time domains under physically meaningful norms, our approach goes beyond the traditional supervised learning framework by effectively leveraging \emph{auxiliary data}---samples of the source function of the heat equation---to design a \emph{physics-informed energy regularization}. Our analysis reveals that learning in linear parabolic PDEs exhibits intrinsic anisotropy between spatial and temporal regularity, in sharp contrast to the static elliptic problems, leading to fundamentally different estimation and generalization behaviors. Specifically, we derive non-asymptotically rate-optimal estimation errors, as well as generalization error, which explicitly capture the interplay among temporal sampling, spatial resolution, and solution smoothness. Our numerical experiments demonstrate substantial orders of magnitude of reduction in testing errors against various physics-informed neural network methods on various parabolic and hyperbolic PDEs.