Hyper-parameter tuning of physics-informed neural networks: Application to Helmholtz problems

Escapil-Inchauspe, Paul; Ruz, Gonzalo A.

Abstract

We consider physics-informed neural networks (PINNs) (Raissiet al., 2019) for forward physical problems. In order to find optimal PINNs configuration, we introduce a hyper-parameter optimization (HPO) procedure via Gaussian processes-based Bayesian optimization. We apply the HPO to Helmholtz equation for bounded domains and conduct a thorough study, focusing on: (i) performance, (ii) the collocation points density r and (iii) the frequency kappa, confirming the applicability and necessity of the method. Numerical experiments are performed in two and three dimensions, including comparison to finite element methods.

Más información

Título según WOS: Hyper-parameter tuning of physics-informed neural networks: Application to Helmholtz problems
Título según SCOPUS: ID SCOPUS_ID:85173236600 Not found in local SCOPUS DB
Título de la Revista: NEUROCOMPUTING
Volumen: 561
Editorial: Elsevier
Fecha de publicación: 2023
DOI:

10.1016/J.NEUCOM.2023.126826

Notas: ISI, SCOPUS