Improving Robustness of Graph Neural Networks with Heterophily-Inspired Designs.
Jiong ZhuJunchen JinMichael T. SchaubDanai KoutraPublished in: CoRR (2021)
Keyphrases
- neural network
- pattern recognition
- fuzzy logic
- artificial neural networks
- graph model
- directed graph
- graph theory
- genetic algorithm
- structured data
- self organizing maps
- computational efficiency
- graph matching
- stable set
- neural nets
- bipartite graph
- design principles
- graph structure
- fuzzy systems
- graph mining
- recurrent neural networks
- graph representation
- adjacency matrix
- fault diagnosis