PDE-Net 2.0: learning PDEs from data with a numeric-symbolic hybrid deep network

DOI10.1016/j.jcp.2019.108925zbMath1454.65131arXiv1812.04426OpenAlexW2903660960WikidataQ114163509 ScholiaQ114163509MaRDI QIDQ2222627

Yiping Lu, Bin Dong, Zichao Long

Publication date: 27 January 2021

Published in: Journal of Computational Physics (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1812.04426

zbMATH Keywords

partial differential equationsdynamic systemconvolutional neural networksymbolic neural network

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Finite difference methods for initial value and initial-boundary value problems involving PDEs (65M06) Numerical methods for partial differential equations, initial value and time-dependent initial-boundary value problems (65M99)

Related Items

Using deep learning to extend the range of air pollution monitoring and forecasting, On some neural network architectures that can represent viscosity solutions of certain high dimensional Hamilton-Jacobi partial differential equations, DLGA-PDE: discovery of PDEs with incomplete candidate library via combination of deep learning and genetic algorithm, PhyGeoNet: physics-informed geometry-adaptive convolutional neural networks for solving parameterized steady-state PDEs on irregular domain, PFNN: a penalty-free neural network method for solving a class of second-order boundary-value problems on complex geometries, Physics-Driven Learning of the Steady Navier-Stokes Equations using Deep Convolutional Neural Networks, Solving inverse-PDE problems with physics-aware neural networks, Image inversion and uncertainty quantification for constitutive laws of pattern formation, DeepM\&Mnet: inferring the electroconvection multiphysics fields based on operator approximation by neural networks, Weak SINDy for partial differential equations, Extended dynamic mode decomposition for inhomogeneous problems, PFNN-2: A Domain Decomposed Penalty-Free Neural Network Method for Solving Partial Differential Equations, Deep-learning based discovery of partial differential equations in integral form from sparse and noisy data, Hybrid FEM-NN models: combining artificial neural networks with the finite element method, A robust framework for identification of PDEs from noisy data, Mesh-Conv: convolution operator with mesh resolution independence for flow field modeling, DeLISA: deep learning based iteration scheme approximation for solving PDEs, Learning time-dependent PDEs with a linear and nonlinear separate convolutional neural network, Revealing hidden dynamics from time-series data by ODENet, A deep neural network-based numerical method for solving contact problems, Robust Identification of Differential Equations by Numerical Techniques from a Single Set of Noisy Observation, Physics-informed PointNet: a deep learning solver for steady-state incompressible flows and thermal fields on multiple sets of irregular geometries, Designing rotationally invariant neural networks from PDEs and variational methods, Structure preservation for the deep neural network multigrid solver, The Discovery of Dynamics via Linear Multistep Methods and Deep Learning: Error Estimation, DeepParticle: learning invariant measure by a deep neural network minimizing Wasserstein distance on data generated from an interacting particle method, NeuralPDE: modelling dynamical systems from data, Data-driven sparse identification of nonlinear dynamical systems using linear multistep methods, Deep learning neural networks for the third-order nonlinear Schrödinger equation: bright solitons, breathers, and rogue waves, The Inverse Heat Transfer Problem of Malan Loess Based on Machine Learning with Finite Element Solver as the Trainer, Simultaneous neural network approximation for smooth functions, PhySR: physics-informed deep super-resolution for spatiotemporal data, Physics-incorporated convolutional recurrent neural networks for source identification and forecasting of dynamical systems, Randomized neural network with Petrov-Galerkin methods for solving linear and nonlinear partial differential equations, Group projected subspace pursuit for identification of variable coefficient differential equations (GP-IDENT), Model discovery of compartmental models with graph-supported neural networks, A Bayesian framework for learning governing partial differential equation from data, Deep-OSG: deep learning of operators in semigroup, Deep Ritz method with adaptive quadrature for linear elasticity, A framework based on symbolic regression coupled with eXtended physics-informed neural networks for gray-box learning of equations of motion from data, Data-driven, multi-moment fluid modeling of Landau damping, Accuracy and architecture studies of residual neural network method for ordinary differential equations, A method for computing inverse parametric PDE problems with random-weight neural networks, Pseudo-Hamiltonian neural networks for learning partial differential equations, Pseudo-Hamiltonian system identification, Transfer learning based physics-informed neural networks for solving inverse problems in engineering structures under different loading scenarios, Neural network architectures using min-plus algebra for solving certain high-dimensional optimal control problems and Hamilton-Jacobi PDEs, Transferable neural networks for partial differential equations, Forward to the special topic on ``Solving differential equations with deep learning, A Review of Data‐Driven Discovery for Dynamic Systems, A Variational Neural Network Approach for Glacier Modelling with Nonlinear Rheology, Deep neural networks learning forward and inverse problems of two-dimensional nonlinear wave equations with rational solitons, Learning the nonlinear flux function of a hidden scalar conservation law from data, Quadrature rule based discovery of dynamics by data-driven denoising, Data-driven learning of differential equations: combining data and model uncertainty, Connections between numerical algorithms for PDEs and neural networks, Identification of the flux function of nonlinear conservation laws with variable parameters, \(\mathrm{U}^p\)-net: a generic deep learning-based time stepper for parameterized spatio-temporal dynamics, A priori error estimate of deep mixed residual method for elliptic PDEs, Is the neural tangent kernel of PINNs deep learning general partial differential equations always convergent?, On mathematical modeling in image reconstruction and beyond, Towards discovery of the differential equations, An overview on deep learning-based approximation methods for partial differential equations, Forecasting of nonlinear dynamics based on symbolic invariance, A review on deep learning in medical image reconstruction, Convolutional Neural Networks in Phase Space and Inverse Problems, Data-driven identification of 2D partial differential equations using extracted physical features, Discovery of Dynamics Using Linear Multistep Methods, A physics-guided neural network framework for elastic plates: comparison of governing equations-based and energy-based approaches, Networks for nonlinear diffusion problems in imaging, Optimization algorithms for parameter identification in parabolic partial differential equations, Artificial neural network approximations of Cauchy inverse problem for linear PDEs, DL-PDE: Deep-Learning Based Data-Driven Discovery of Partial Differential Equations from Discrete and Noisy Data, Enforcing Imprecise Constraints on Generative Adversarial Networks for Emulating Physical Systems, A physics-informed operator regression framework for extracting data-driven continuum models, Continuous-time system identification with neural networks: model structures and fitting criteria, Frame-independent vector-cloud neural network for nonlocal constitutive modeling on arbitrary grids, PDE-Net, Jets and differential linear logic, Bayesian learning of stochastic dynamical models, Data-Driven Learning of Nonautonomous Systems, Overcoming the curse of dimensionality for some Hamilton-Jacobi partial differential equations via neural network architectures, A Tailored Convolutional Neural Network for Nonlinear Manifold Learning of Computational Physics Data Using Unstructured Spatial Discretizations, A Bayesian approach for data-driven dynamic equation discovery, Latent-space time evolution of non-intrusive reduced-order models using Gaussian process emulation

Uses Software

• torchdiffeq

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Discovering governing equations from data by sparse identification of nonlinear dynamical systems
• Weighted essentially non-oscillatory schemes
• Hidden physics models: machine learning of nonlinear partial differential equations
• Provable approximation properties for deep neural networks
• Exponential convergence of the deep neural network approximation for analytic functions
• Learning physics by data for the motion of a sphere falling in a non-Newtonian fluid
• Identification of physical processes via combined data-driven and data-assimilation methods
• Data driven governing equations approximation using deep neural networks
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• A proposal on machine learning via dynamical systems
• Relu Deep Neural Networks and Linear Finite Elements
• Automated reverse engineering of nonlinear dynamical systems
• Ten Lectures on Wavelets
• Finite Volume Methods for Hyperbolic Problems
• Mathematical Modeling of Mitochondrial Swelling
• Stable architectures for deep neural networks
• Learning partial differential equations via data discovery and sparse optimization
• Extracting Structured Dynamical Systems Using Sparse Optimization With Very Few Samples
• Model Reduction with Memory and the Machine Learning of Dynamical Systems
• Image restoration: Total variation, wavelet frames, and beyond
• Turbulence Modeling in the Age of Data
• Image Restoration: Wavelet Frame Shrinkage, Nonlinear Evolution PDEs, and Beyond
• Deep learning for physical processes: incorporating prior scientific knowledge