Gauss Newton method for solving variational problems of PDEs with neural network discretizaitons
DOI10.1007/S10915-024-02535-ZMaRDI QIDQ6569679FDOQ6569679
Authors: Wenrui Hao, Qingguo Hong, Xianlin Jin 
Publication date: 9 July 2024
Published in: Journal of Scientific Computing (Search for Journal in Brave)
convergence analysispartial differential equationsGauss-Newton methodvariational formneural network discretization
Monte Carlo methods (65C05) Numerical optimization and variational techniques (65K10) Artificial neural networks and deep learning (68T07) Eigenvalues, singular values, and eigenvectors (15A18) Numerical solutions to overdetermined systems, pseudoinverses (65F20) Computational learning theory (68Q32) Numerical computation of solutions to systems of equations (65H10) Variational methods applied to PDEs (35A15) Integro-partial differential equations (35R09) Series expansions (e.g., Taylor, Lidstone series, but not Fourier series) (41A58) Stability and convergence of numerical methods for boundary value problems involving PDEs (65N12) Second-order elliptic equations (35J15) PDEs on graphs and networks (ramified or polygonal spaces) (35R02) Numerical methods for partial differential equations, boundary value problems (65N99)
Cites Work
- DGM: a deep learning algorithm for solving partial differential equations
- fPINNs: Fractional Physics-Informed Neural Networks
- Universal approximation bounds for superpositions of a sigmoidal function
- Numerically solving polynomial systems with Bertini
- Note on the Generalized Inverse of a Matrix Product
- Multiplicative perturbation theory of the Moore-Penrose inverse and the least squares problem
- Perturbation theory for pseudo-inverses
- Approximation rates for neural networks with general activation functions
- MONTE CARLO METHODS FOR SOLVING MULTIVARIABLE PROBLEMS
- Neural‐network‐based approximations for solving partial differential equations
- Solving parametric PDE problems with artificial neural networks
- Solving high-dimensional partial differential equations using deep learning
- The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- The Stone-Weierstrass Theorem
- Randomized Newton's method for solving differential equations based on the neural network discretization
- Gradient descent optimizes over-parameterized deep ReLU networks
- ReLU deep neural networks and linear finite elements
- Deep Network Approximation for Smooth Functions
- Approximation by Combinations of ReLU and Squared ReLU Ridge Functions With <inline-formula> <tex-math notation="LaTeX">$\ell^1$ </tex-math> </inline-formula> and <inline-formula> <tex-math notation="LaTeX">$\ell^0$ </tex-math> </inline-formula> Controls
- Finite neuron method and convergence analysis
- Optimal approximation rate of ReLU networks in terms of width and depth
- High-order approximation rates for shallow neural networks with cosine and \(\mathrm{ReLU}^k\) activation functions
- SelectNet: self-paced learning for high-dimensional partial differential equations
- HomPINNs: Homotopy physics-informed neural networks for learning multiple solutions of nonlinear elliptic differential equations
- A Newton’s iteration converges quadratically to nonisolated solutions too
- Greedy training algorithms for neural networks and applications to PDEs
- A priori generalization error analysis of two-layer neural networks for solving high dimensional Schrödinger eigenvalue problems
This page was built for publication: Gauss Newton method for solving variational problems of PDEs with neural network discretizaitons
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6569679)
