Communication Dans Un Congrès Année : 2025

Provably Accurate Adaptive Sampling for Collocation Points in Physics-informed Neural Networks

Résumé

Despite considerable scientific advances in numerical simulation, efficiently solving PDEs remains a complex and often expensive problem. Physics-informed Neural Networks (PINN) have emerged as an efficient way to learn surrogate solvers by embedding the PDE in the loss function and minimizing its residuals using automatic differentiation at so-called collocation points. Originally uniformly sampled, the choice of the latter has been the subject of recent advances leading to adaptive sampling refinements for PINNs. In this paper, leveraging a new quadrature method for approximating definite integrals, we introduce a provably accurate sampling method for collocation points based on the Hessian of the PDE residuals. Comparative experiments conducted on a set of 1D and 2D PDEs demonstrate the benefits of our method.

Fichier principal
Vignette du fichier
Provably accurate adaptive sampling for collocation points in PINN.pdf (1.31 Mo) Télécharger le fichier

Dates et versions

hal-05017201 , version 1 (02-04-2025)
hal-05017201 , version 2 (04-04-2025)
hal-05017201 , version 3 (10-06-2025)

Licence

Identifiants

Citer

Antoine Caradot, Rémi Emonet, Abdel-Rahim Mezidi, Amaury Habrard, Marc Sebban. Provably Accurate Adaptive Sampling for Collocation Points in Physics-informed Neural Networks. European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Sep 2025, Porto, Portugal. ⟨hal-05017201v2⟩
746 Consultations
579 Téléchargements

Altmetric

Partager

  • More