On Explainability of Graph Neural Networks via Subgraph Explorations.
Hao YuanHaiyang YuJie WangKang LiShuiwang JiPublished in: CoRR (2021)
Keyphrases
- neural network
- graph mining
- graph databases
- graph properties
- graph data
- graph classification
- subgraph isomorphism
- labeled graphs
- maximum weight
- graph representation
- subgraph mining
- query graph
- structured data
- maximum matching
- frequent subgraph mining
- artificial neural networks
- graph theory
- similarity graph
- graph matching
- recurrent neural networks
- graph structure
- genetic algorithm
- edge weights
- graph search
- connected subgraphs
- pattern recognition
- graph model
- random walk
- relational data
- directed graph
- self organizing maps
- graph theoretic
- subgraph matching
- back propagation
- maximum clique
- pattern mining
- frequent subgraphs
- graph partitioning
- fault diagnosis
- stable set
- graph structures
- multi layer
- sufficient conditions
- fuzzy logic
- activation function
- image segmentation