Multipole Graph Neural Operator for Parametric Partial Differential Equations

Abstract

One of the main challenges in using deep learning-based methods for simulating physical systems and solving partial differential equations (PDEs) is formulating physics-based data in the desired structure for neural networks. Graph neural networks (GNNs) have gained popularity in this area since graphs offer a natural way of modeling particle interactions and provide a clear way of discretizing the continuum models. However, the graphs constructed for approximating such tasks usually ignore long-range interactions due to unfavorable scaling of the computational complexity with respect to the number of nodes. The errors due to these approximations scale with the discretization of the system, thereby not allowing for generalization under mesh-refinement. Inspired by the classical multipole methods, we propose a novel multi-level graph neural network framework that captures interaction at all ranges with only linear complexity. Our multi-level formulation is equivalent to recursively adding inducing points to the kernel matrix, unifying GNNs with multi-resolution matrix factorization of the kernel. Experiments confirm our multi-graph network learns discretization-invariant solution operators to PDEs and can be evaluated in linear time.

References in corpus (18)

• Semi-Supervised Classification with Graph Convolutional Networks
• Inductive Representation Learning on Large Graphs
• Neural Message Passing for Quantum Chemistry
• Fast Graph Representation Learning with PyTorch Geometric
• Bayesian Deep Convolutional Encoder-Decoder Networks for Surrogate Modeling and Uncertainty Quantification
• MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing
• EikoNet: Solving the Eikonal equation with Deep Neural Networks
• Gauge Equivariant Convolutional Networks and the Icosahedral CNN
• Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids
• Flexible Neural Representation for Physics Prediction
• Neural Operator: Graph Kernel Network for Partial Differential Equations
• The Random Feature Model for Input-Output Maps between Banach Spaces
• MgNet: A Unified Framework of Multigrid and Convolutional Neural Network
• N-GCN: Multi-scale Graph Convolution for Semi-supervised Node Classification
• Janossy Pooling: Learning Deep Permutation-Invariant Functions for Variable-Size Inputs
• Gauge Equivariant Mesh CNNs: Anisotropic convolutions on geometric graphs
• Dynamic Multiscale Graph Neural Networks for 3D Skeleton-Based Human Motion Prediction
• Unsupervised Deep Learning Algorithm for PDE-based Forward and Inverse Problems

Cited by in corpus (31)

• Enhancing Computational Fluid Dynamics with Machine Learning
• A physics-informed variational DeepONet for predicting the crack path in brittle materials
• Physics-informed graph neural Galerkin networks: A unified framework for solving PDE-governed forward and inverse problems
• Real-time high-resolution CO${}_2$ geological storage prediction using nested Fourier neural operators
• Interfacing Finite Elements with Deep Neural Operators for Fast Multiscale Modeling of Mechanics Problems
• Choose a Transformer: Fourier or Galerkin
• Deep Learning Methods for Partial Differential Equations and Related Parameter Identification Problems
• Learning macroscopic internal variables and history dependence from microscopic models
• Nonlocal Kernel Network (NKN): a Stable and Resolution-Independent Deep Neural Network
• On universal approximation and error bounds for Fourier Neural Operators
• Residual-based error correction for neural operator accelerated infinite-dimensional Bayesian inverse problems
• Artificial Intelligence in Materials Science and Engineering: Current Landscape, Key Challenges, and Future Trajectorie
• Applications of Scientific Machine Learning for the Analysis of Functionally Graded Porous Beams
• Solving and Learning Nonlinear PDEs with Gaussian Processes
• Magnetohydrodynamics with Physics Informed Neural Operators
• Fourier neural operator for learning solutions to macroscopic traffic flow models: Application to the forward and inverse problems
• Meta-Auto-Decoder for Solving Parametric Partial Differential Equations
• Deep Learning of Conjugate Mappings
• Neural Dynamical Operator: Continuous Spatial-Temporal Model with Gradient-Based and Derivative-Free Optimization Methods
• Continuous PDE Dynamics Forecasting with Implicit Neural Representations
• A Priori Analysis of Stable Neural Network Solutions to Numerical PDEs
• Predicting Change, Not States: An Alternate Framework for Neural PDE Surrogates
• Frame-independent vector-cloud neural network for nonlocal constitutive modeling on arbitrary grids
• Inverting the Kohn-Sham equations with physics-informed machine learning
• Seismic wave propagation and inversion with Neural Operators
• Frame invariance and scalability of neural operators for partial differential equations
• MCMC-Net: Accelerating Markov Chain Monte Carlo with Neural Networks for Inverse Problems
• Efficient Long-Range Convolutions for Point Clouds
• SF-SGL: Solver-Free Spectral Graph Learning from Linear Measurements
• Approximation of Functionals by Neural Network without Curse of Dimensionality
• FC2T2: The Fast Continuous Convolutional Taylor Transform with Applications in Vision and Graphics