A Deep Double Ritz Method (DRM) for solving Partial Differential Equations using Neural Networks
arXiv:2211.03627 · doi:10.1016/j.cma.2023.115892
Abstract
Residual minimization is a widely used technique for solving Partial Differential Equations in variational form. It minimizes the dual norm of the residual, which naturally yields a saddle-point (min-max) problem over the so-called trial and test spaces. In the context of neural networks, we can address this min-max approach by employing one network to seek the trial minimum, while another network seeks the test maximizers. However, the resulting method is numerically unstable as we approach the trial solution. To overcome this, we reformulate the residual minimization as an equivalent minimization of a Ritz functional fed by optimal test functions computed from another Ritz functional minimization. We call the resulting scheme the Deep Double Ritz Method (DRM), which combines two neural networks for approximating trial functions and optimal test functions along a nested double Ritz minimization strategy. Numerical results on different diffusion and convection problems support the robustness of our method, up to the approximation properties of the networks and the training capacity of the optimizers.
arXiv comment28 pages
References in corpus (21)
- Adam: A Method for Stochastic Optimization
- TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
- An overview of gradient descent optimization algorithms
- DGM: A deep learning algorithm for solving partial differential equations
- hp-VPINNs: Variational Physics-Informed Neural Networks With Domain Decomposition
- Weak Adversarial Networks for High-dimensional Partial Differential Equations
- A Review of automatic differentiation and its efficient implementation
- Variational Physics-Informed Neural Networks For Solving Partial Differential Equations
- Introduction to Monte Carlo methods
- Numerical Solution of Inverse Problems by Weak Adversarial Networks
- On quadrature rules for solving Partial Differential Equations using Neural Networks
- An adaptive stabilized conforming finite element method via residual minimization on dual discontinuous Galerkin norms
- Data-Driven Finite Elements Methods: Machine Learning Acceleration of Goal-Oriented Computations
- (Parametrized) First Order Transport Equations: Realization of Optimally Stable Petrov-Galerkin Methods
- An automatic-adaptivity stabilized finite element method via residual minimization for heterogeneous, anisotropic advection-diffusion-reaction problems
- Non-Euclidean Universal Approximation
- Five lectures on DPG methods
- Isogeometric Residual Minimization Method (iGRM) with Direction Splitting Preconditoner for Stationary Advection-Diffusion Problems
- The DPG-star method
- Neural Control of Discrete Weak Formulations: Galerkin, Least-Squares and Minimal-Residual Methods with Quasi-Optimal Weights
- Deep Petrov-Galerkin Method for Solving Partial Differential Equations