A priori and a posteriori error estimates for the deep Ritz method applied to the Laplace and Stokes problem

DOI10.1016/J.CAM.2022.114845zbMATH Open1502.65213arXiv2107.11035OpenAlexW4298111905MaRDI QIDQ2095152FDOQ2095152

Authors: Yanyan Li 

Publication date: 9 November 2022

Published in: Journal of Computational and Applied Mathematics (Search for Journal in Brave)

Abstract: We analyze neural network solutions to partial differential equations obtained with Physics Informed Neural Networks. In particular, we apply tools of classical finite element error analysis to obtain conclusions about the error of the Deep Ritz method applied to the Laplace and the Stokes equations. Further, we develop an a posteriori error estimator for neural network approximations of partial differential equations. The proposed approach is based on the dual weighted residual estimator. It is destined to serve as a stopping criterion that guarantees the accuracy of the solution independently of the design of the neural network training. The result is equipped with computational examples for Laplace and Stokes problems.

Full work available at URL: https://arxiv.org/abs/2107.11035

Recommendations

• Solving PDEs by variational physics-informed neural networks: an a posteriori error analysis
• Convergence rate analysis for deep Ritz method
• Error estimates for deep learning methods in fluid dynamics
• The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems
• Deep Nitsche Method: Deep Ritz Method with Essential Boundary Conditions

zbMATH Keywords

a posteriori error estimateserror estimatesfinite elementsneural networksdual weighted residual method

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Laplace operator, Helmholtz equation (reduced wave equation), Poisson equation (35J05) Error bounds for boundary value problems involving PDEs (65N15) Finite element, Rayleigh-Ritz and Galerkin methods for boundary value problems involving PDEs (65N30) Stokes and related (Oseen, etc.) flows (76D07) Finite element methods applied to problems in fluid mechanics (76M10)

Cites Work

• DGM: a deep learning algorithm for solving partial differential equations
• Universal approximation bounds for superpositions of a sigmoidal function
• Title not available (Why is that?)
• An optimal control approach to a posteriori error estimation in finite element methods
• Title not available (Why is that?)
• A posteriori error control in finite element methods via duality techniques: Application to perfect plasticity
• Variational localizations of the dual weighted residual estimator
• A posterior error estimation for finite element discretization of parameter identification problems
• Goal-oriented space-time adaptivity in the finite element Galerkin method for the computation of nonstationary incompressible flow
• Finite element methods for incompressible flow problems
• The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems
• A machine learning framework for data driven acceleration of computations of differential equations
• A posteriori error estimation for the fractional step theta discretization of the incompressible Navier-Stokes equations
• Fluid-structure interactions. Models, analysis and finite elements
• DeepXDE: a deep learning library for solving differential equations
• An energy approach to the solution of partial differential equations in computational mechanics via machine learning: concepts, implementation and applications
• A neural network multigrid solver for the Navier-Stokes equations
• Deep Nitsche Method: Deep Ritz Method with Essential Boundary Conditions
• Structure preservation for the deep neural network multigrid solver
• Estimates on the generalization error of physics-informed neural networks for approximating a class of inverse problems for PDEs

Cited In (14)

• Convergence rate analysis for deep Ritz method
• Error analysis of deep Ritz methods for elliptic equations
• Generalization error in the deep Ritz method with smooth activation functions
• Inf-sup neural networks for high-dimensional elliptic PDE problems
• Phase field smoothing-PINN: a neural network solver for partial differential equations with discontinuous coefficients
• Posteriori error neural network: a recovery type posteriori error estimator based on neural network for diffusion problems
• Error estimates for deep learning methods in fluid dynamics
• Deep Nitsche Method: Deep Ritz Method with Essential Boundary Conditions
• DNN-MG: a hybrid neural network/finite element method with applications to 3D simulations of the Navier-Stokes equations
• Optimal a priori error bounds for the Rayleigh-Ritz method
• Improving weak PINNs for hyperbolic conservation laws: dual norm computation, boundary conditions and systems
• A comparison study of deep Galerkin method and deep Ritz method for elliptic problems with different boundary conditions
• Deep Ritz method with adaptive quadrature for linear elasticity
• Solving PDEs by variational physics-informed neural networks: an a posteriori error analysis

Uses Software

• GASCOIGNE
• PyTorch
• Adam
• DeepXDE
• DGM
• DeepONet

This page was built for publication: A priori and a posteriori error estimates for the deep Ritz method applied to the Laplace and Stokes problem

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2095152)