Analysis of Deep Ritz Methods for Laplace Equations with Dirichlet Boundary Conditions

Abstract

Deep Ritz methods (DRM) have been proven numerically to be efficient in solving partial differential equations. In this paper, we present a convergence rate in $H^1$ norm for deep Ritz methods for Laplace equations with Dirichlet boundary condition, where the error depends on the depth and width in the deep neural networks and the number of samples explicitly. Further we can properly choose the depth and width in the deep neural networks in terms of the number of training samples. The main idea of the proof is to decompose the total error of DRM into three parts, that is approximation error, statistical error and the error caused by the boundary penalty. We bound the approximation error in $H^1$ norm with ${\mathrm{ReLU}}^2$ networks and control the statistical error via Rademacher complexity. In particular, we derive the bound on the Rademacher complexity of the non-Lipschitz composition of gradient norm with ${\mathrm{ReLU}}^2$ network, which is of immense independent interest. We also analysis the error inducing by the boundary penalty method and give a prior rule for tuning the penalty parameter.

arXiv commentarXiv admin note: substantial text overlap with arXiv:2103.13330; text overlap with arXiv:2109.01780

References in corpus (12)

• DGM: A deep learning algorithm for solving partial differential equations
• Solving high-dimensional partial differential equations using deep learning
• Weak Adversarial Networks for High-dimensional Partial Differential Equations
• Nearly-tight VC-dimension and pseudodimension bounds for piecewise linear neural networks
• When and why PINNs fail to train: A neural tangent kernel perspective
• The Finite Neuron Method and Convergence Analysis
• Learning Neural PDE Solvers with Convergence Guarantees
• Two-Layer Neural Networks for Partial Differential Equations: Optimization and Generalization Theory
• Error estimates of residual minimization using neural networks for linear PDEs
• Estimates on the generalization error of Physics Informed Neural Networks (PINNs) for approximating PDEs
• Learning to Optimize Multigrid PDE Solvers
• A Priori Generalization Analysis of the Deep Ritz Method for Solving High Dimensional Elliptic Equations

Cited by in corpus (1)

• Certified machine learning: Rigorous a posteriori error bounds for PDE defined PINNs