Convergence of Physics-Informed Neural Networks Applied to Linear Second-Order Elliptic Interface Problems

DOI10.4208/cicp.OA-2022-0218MaRDI QIDQ5887902

Benzhuo Lu, Aiqing Zhu, Sidi Wu, Yi-Fa Tang

Publication date: 21 April 2023

Published in: Communications in Computational Physics (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2203.03407

zbMATH Keywords

convergence analysisneural networkselliptic interface problemsgeneralization errors

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Stability and convergence of numerical methods for boundary value problems involving PDEs (65N12) Interface problems; diffusion-limited aggregation arising in equilibrium statistical mechanics (82B24)

Related Items (3)

A Rate of Convergence of Weak Adversarial Neural Networks for the Second Order Parabolic PDEs ⋮ HRW: Hybrid Residual and Weak Form Loss for Solving Elliptic Interface Problems with Neural Network ⋮ Less Emphasis on Hard Regions: Curriculum Learning of PINNs for Singularly Perturbed Convection-Diffusion-Reaction Problems

Uses Software

• DiffSharp
• Adam
• FPINNs

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations
• A numerical method for solving variable coefficient elliptic equation with interfaces
• Weak adversarial networks for high-dimensional partial differential equations
• A bilinear partially penalized immersed finite element method for elliptic interface problems with multi-domain and triple-junction points
• A hybrid method for moving interface problems with application to the Hele-Shaw flow
• Approximation rates for neural networks with encodable weights in smoothness spaces
• INN: interfaced neural networks as an accessible meshless approach for solving interface PDE problems
• On some neural network architectures that can represent viscosity solutions of certain high dimensional Hamilton-Jacobi partial differential equations
• Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems
• A Petrov-Galerkin finite element interface method for interface problems with Bloch-periodic boundary conditions and its application in phononic crystals
• Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
• A mesh-free method for interface problems using the deep learning approach
• A moment-of-fluid method for diffusion equations on irregular domains in multi-material systems
• Efficient multigrid solution of elliptic interface problems using viscosity-upwinded local discontinuous Galerkin methods
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• Solving elliptic problems with discontinuities on irregular domains -- the Voronoi interface method
• The finite element method for elliptic equations with discontinuous coefficients
• Neural algorithm for solving differential equations
• The immersed boundary method
• On the poisson equation with intersecting interfaces
• Universal approximation bounds for superpositions of a sigmoidal function
• The Immersed Interface Method for Elliptic Equations with Discontinuous Coefficients and Singular Sources
• An Unfitted Discontinuous Galerkin Method Applied to Elliptic Interface Problems
• Consistency of Lipschitz Learning with Infinite Unlabeled Data and Finite Labeled Data
• Analysis of the Generalization Error: Empirical Risk Minimization over Deep Artificial Neural Networks Overcomes the Curse of Dimensionality in the Numerical Approximation of Black--Scholes Partial Differential Equations
• Extended Physics-Informed Neural Networks (XPINNs): A Generalized Space-Time Domain Decomposition Based Deep Learning Framework for Nonlinear Partial Differential Equations
• On the Convergence of Physics Informed Neural Networks for Linear Second-Order Elliptic and Parabolic Type PDEs
• fPINNs: Fractional Physics-Informed Neural Networks
• Approximation by superpositions of a sigmoidal function
• Neural network approximation: three hidden layers are enough
• On the approximation of functions by tanh neural networks

This page was built for publication: Convergence of Physics-Informed Neural Networks Applied to Linear Second-Order Elliptic Interface Problems