Hyper-SAGNN: a self-attention based graph neural network for hypergraphs.
Ruochi ZhangYuesong ZouJian MaPublished in: ICLR (2020)
Keyphrases
- neural network
- graph theory
- random walk
- hypergraph model
- neural network model
- pattern recognition
- artificial neural networks
- bipartite graph
- graph model
- graph structure
- network model
- higher order relations
- structured data
- graph representation
- graph clustering
- learning vector quantization
- neural network is trained
- fuzzy logic
- directed graph
- neural nets
- knn
- associative memory
- weighted graph
- prediction model
- multilayer perceptron
- activation function
- feed forward neural networks
- dependency graph
- auto associative