Physics-Embedded Neural Networks: E(n)-Equivariant Graph Neural PDE Solvers.
Masanobu HorieNaoto MitsumePublished in: CoRR (2022)
Keyphrases
- neural network
- network architecture
- matrix valued
- neural model
- associative memory
- partial differential equations
- artificial neural networks
- graph representation
- pattern recognition
- graph structure
- neural learning
- connectionist models
- vector valued
- structured data
- neural fuzzy
- genetic algorithm
- learning rules
- computer science
- embedded systems
- high order
- bipartite graph
- artificial intelligence
- fuzzy logic
- graph theory
- recurrent neural networks
- multilayer perceptron
- neural network model
- dt mri
- neural computation
- bio inspired
- fuzzy systems
- weighted graph
- sat solvers
- graph matching
- directed graph
- random walk
- graph model
- graph databases
- rotation invariant
- anisotropic diffusion
- scale spaces
- feed forward
- self organizing maps
- level set