PhyGNNet: Solving spatiotemporal PDEs with Physics-informed Graph Neural Network.
Longxiang JiangLiyuan WangXinkun ChuYonghao XiaoHao ZhangPublished in: CoRR (2022)
Keyphrases
- neural network
- graph model
- computer science
- artificial neural networks
- random walk
- back propagation
- space time
- structured data
- multiscale
- graph structure
- connected components
- directed graph
- graph theory
- bipartite graph
- genetic algorithm
- network architecture
- neural network model
- partial differential equations
- artificial intelligence
- graph matching
- prediction model
- neural nets
- combinatorial optimization
- directed acyclic graph
- knn
- graph representation
- graph theoretic
- numerical solution