Cover Image

Are Deep Learning Based Hybrid PDE Solvers Reliable? Why Training Paradigms and Update Strategies Matter

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Wu, Yuhan, van Beek, Jan Willem, Dolean, Victorita, Heinlein, Alexander
Format:	Preprint
Published:	2026
Subjects:	Numerical Analysis Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Local Feature Filtering for Scalable and Well-Conditioned Domain-Decomposed Random Feature Methods
by: van Beek, Jan Willem, et al.
Published: (2025)

Multilevel domain decomposition-based architectures for physics-informed neural networks
by: Dolean, Victorita, et al.
Published: (2023)

Nonlinear Multilevel Solution Strategies for Diffusive Wave Flood Models in Perforated Domains
by: Boutilier, Miranda, et al.
Published: (2024)

Robust Methods for Multiscale Coarse Approximations of Diffusion Models in Perforated Domains
by: Boutilier, Miranda, et al.
Published: (2023)

Preconditioning of GMRES for Helmholtz problems with quasimodes
by: Dolean, Victorita, et al.
Published: (2025)

ELM-FBPINNs: An Efficient Multilevel Random Feature Method
by: Anderson, Samuel, et al.
Published: (2024)

Modal analysis of a domain decomposition method for Maxwell's equations in a waveguide
by: Dolean, Victorita, et al.
Published: (2025)

Numerical assessment of PML transmission conditions in a domain decomposition method for the Helmholtz equation
by: Bootland, Niall, et al.
Published: (2022)

Neural network-driven domain decomposition for efficient solutions to the Helmholtz equation
by: Dolean, Victorita, et al.
Published: (2025)

A Guided Tour of Modern Domain Decomposition: From Schwarz Iterations to Robust Preconditioners and HPC Implementations
by: Dolean, Victorita, et al.
Published: (2026)

A two-level domain-decomposition preconditioner for the time-harmonic Maxwell's equations
by: Bonazzoli, Marcella, et al.
Published: (2017)

Two-level preconditioners for the Helmholtz equation
by: Bonazzoli, Marcella, et al.
Published: (2017)

Spectral coarse spaces based on indefinite operators: the $H_k$-GenEO method
by: Chaumont-Frelet, Théophile, et al.
Published: (2026)

Can Symmetric Positive Definite (SPD) coarse spaces perform well for indefinite Helmholtz problems?
by: Dolean, Victorita, et al.
Published: (2024)

Nodal Hybrid Neural Solvers for Parametric PDE Systems
by: Liu, Yun, et al.
Published: (2025)

Stable Weight Updating: A Key to Reliable PDE Solutions Using Deep Learning
by: Noorizadegan, A., et al.
Published: (2024)

An Iterative Block Matrix Inversion (IBMI) Algorithm for Symmetric Positive Definite Matrices with Applications to Covariance Matrices
by: Paterson, Ann, et al.
Published: (2025)

Convergence Guarantees for Gradient-Based Training of Neural PDE Solvers: From Linear to Nonlinear PDEs
by: Zhao, Wei, et al.
Published: (2025)

Quantifying Training Difficulty and Accelerating Convergence in Neural Network-Based PDE Solvers
by: Chen, Chuqi, et al.
Published: (2024)

End-to-End Mesh Optimization of a Hybrid Deep Learning Black-Box PDE Solver
by: Ma, Shaocong, et al.
Published: (2024)

Convergence analysis of GMRES applied to Helmholtz problems near resonances
by: Dolean, Victorita, et al.
Published: (2025)

Schwarz preconditioner with $H_k$-GenEO coarse space for the indefinite Helmholtz problem
by: Dolean, Victorita, et al.
Published: (2024)

When can a neural operator replace a coarse solve? Architectural principles for two-level preconditioning
by: Melchers, Hugo, et al.
Published: (2026)

Learning Neural PDE Solvers with Convergence Guarantees
by: Hsieh, Jun-Ting, et al.
Published: (2019)

Coarse Spaces Based on Higher-Order Interpolation for Schwarz Preconditioners for Helmholtz Problems
by: Sieburgh, Erik, et al.
Published: (2024)

A robust and adaptive GenEO-type domain decomposition preconditioner for $\mathbf{H}(\mathbf{curl})$ problems in three-dimensional general topologies
by: Bootland, Niall, et al.
Published: (2023)

Two-level additive Schwarz preconditioners for reduced integration methods
by: Cumaru, Filipe, et al.
Published: (2025)

Sharpened PCG Iteration Bound for High-Contrast Heterogeneous Scalar Elliptic PDEs
by: Soliman, Philip, et al.
Published: (2025)

Coupling deal.II and FROSch: A Sustainable and Accessible (O)RAS Preconditioner
by: Heinlein, Alexander, et al.
Published: (2024)

Domain decomposition architectures and Gauss-Newton training for physics-informed neural networks
by: Heinlein, Alexander, et al.
Published: (2025)

Unisolver: PDE-Conditional Transformers Towards Universal Neural PDE Solvers
by: Zhou, Hang, et al.
Published: (2024)

Hybrid PDE-Deep Neural Network Model for Calcium Dynamics in Neurons
by: Gurung, Abel, et al.
Published: (2024)

Achieving wavenumber robustness in domain decomposition for heterogeneous Helmholtz equation: an overview of spectral coarse spaces
by: Dolean, Victorita, et al.
Published: (2025)

Efficient approximation of high-frequency Helmholtz solutions by Gaussian coherent states
by: Chaumont-Frelet, T., et al.
Published: (2022)

Stability in Training PINNs for Stiff PDEs: Why Initial Conditions Matter
by: Hao, Baoli, et al.
Published: (2024)

PDE Solvers Should Be Local: Fast, Stable Rollouts with Learned Local Stencils
by: Cheng, Chun-Wun, et al.
Published: (2025)

Deep Learning-Enhanced Preconditioning for Efficient Conjugate Gradient Solvers in Large-Scale PDE Systems
by: Li, Rui, et al.
Published: (2024)

Overlapping Schwarz Preconditioners for Randomized Neural Networks with Domain Decomposition
by: Shang, Yong, et al.
Published: (2024)

A computational study of algebraic coarse spaces for two-level overlapping additive Schwarz preconditioners
by: Alves, Filipe A. C. S., et al.
Published: (2024)

Numerical study on hyper parameter settings for neural network approximation to partial differential equations
by: Yang, Hee Jun, et al.
Published: (2025)