Solving parametric partial differential equations with deep rectified quadratic unit neural networks

DOI10.1007/S10915-022-02015-2MaRDI QIDQ2103467zbMATHOpenAlexWikidataFDO

Authors Yanyan Li

Publication date 13 December 2022

Published in Journal of Scientific Computing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2203.06973

zbMATH Keywords

reduced basis methodcomplexity boundsdeep neural networkReQU activation functionsolution maps of parametric PDE

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Finite element, Rayleigh-Ritz and Galerkin methods for boundary value problems involving PDEs (65N30) Numerical methods for partial differential equations, boundary value problems (65N99)

Abstract: Implementing deep neural networks for learning the solution maps of parametric partial differential equations (PDEs) turns out to be more efficient than using many conventional numerical methods. However, limited theoretical analyses have been conducted on this approach. In this study, we investigate the expressive power of deep rectified quadratic unit (ReQU) neural networks for approximating the solution maps of parametric PDEs. The proposed approach is motivated by the recent important work of G. Kutyniok, P. Petersen, M. Raslan and R. Schneider (Gitta Kutyniok, Philipp Petersen, Mones Raslan, and Reinhold Schneider. A theoretical analysis of deep neural networks and parametric pdes. Constructive Approximation, pages 1-53, 2021), which uses deep rectified linear unit (ReLU) neural networks for solving parametric PDEs. In contrast to the previously established complexity-bound $mathcalOleft(d^{3}lo{g}_2^q(1/epsilon)ight)$ for ReLU neural networks, we derive an upper bound $mathcalOleft(d^{3}lo{g}_2^{q}lo{g}_2(1/epsilon)ight)$ on the size of the deep ReQU neural network required to achieve accuracy $epsilon>0$, where $d$ is the dimension of reduced basis representing the solutions. Our method takes full advantage of the inherent low-dimensionality of the solution manifolds and better approximation performance of deep ReQU neural networks. Numerical experiments are performed to verify our theoretical result.

Recommendations

• Numerical solution of the parametric diffusion equation by deep neural networks
• Efficient approximation of solutions of parametric linear transport equations by ReLU DNNs
• Solving parametric PDE problems with artificial neural networks
• A deep learning approach to Reduced Order Modelling of parameter dependent partial differential equations
• Deep learning in high dimension: ReLU neural network expression for Bayesian PDE inversion

Cites work

• scientific article; zbMATH DE number 7626805 (Why is no real title available?)
• scientific article; zbMATH DE number 5681750 (Why is no real title available?)
• A theoretical analysis of deep neural networks and parametric PDEs
• Approximation properties of a multilayered feedforward artificial neural network
• Better approximations of high dimensional smooth functions by deep neural networks with rectified power units
• Certified reduced basis methods for parametrized partial differential equations
• Convergence rate analysis for deep Ritz method
• DGM: a deep learning algorithm for solving partial differential equations
• DNN expression rate analysis of high-dimensional PDEs: application to option pricing
• Data driven approximation of parametrized PDEs by reduced basis and neural networks
• Deep learning in high dimension: neural network expression rates for generalized polynomial chaos expansions in UQ
• Error bounds for approximations with deep ReLU networks
• Error estimates for DeepONets: a deep learning framework in infinite dimensions
• Exponential ReLU DNN expression of holomorphic maps in high dimension
• Kolmogorov widths and low-rank approximations of parametric elliptic PDEs
• Model reduction and neural networks for parametric PDEs
• Non intrusive reduced order modeling of parametrized PDEs by kernel POD and neural networks
• Numerical solution of the parametric diffusion equation by deep neural networks
• ReLU deep neural networks and linear finite elements
• Reduced basis methods for partial differential equations. An introduction
• Solving high-dimensional partial differential equations using deep learning
• Solving parametric PDE problems with artificial neural networks

Cited in

(10)

• Emulated digital CNN-UM solution of partial differential equations
• Deep learning in high dimension: ReLU neural network expression for Bayesian PDE inversion
• Neural network-based variational methods for solving quadratic porous medium equations in high dimensions
• Solving second-order nonlinear evolution partial differential equations using deep learning
• Efficient approximation of solutions of parametric linear transport equations by ReLU DNNs
• A theoretical analysis of deep neural networks and parametric PDEs
• On quadrature rules for solving partial differential equations using neural networks
• Long-time integration of parametric evolution equations with physics-informed DeepONets
• Numerical solution of the parametric diffusion equation by deep neural networks
• scientific article; zbMATH DE number 7546040 (Why is no real title available?)

Describes a project that uses

Uses Software

• FEniCS
• redbKIT
• PyTorch
• DGM
• DeepONet

This page was built for publication: Solving parametric partial differential equations with deep rectified quadratic unit neural networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2103467)