Towards Robust Graph Neural Networks for Noisy Graphs with Sparse Labels.
Enyan DaiWei JinHui LiuSuhang WangPublished in: WSDM (2022)
Keyphrases
- labeled graphs
- neural network
- graph representation
- graph theory
- graph structure
- weighted graph
- label noise
- graph model
- directed graph
- pattern recognition
- gaussian graphical models
- graph databases
- directed acyclic
- edge weights
- graph theoretic
- graph theoretical
- graph construction
- graph matching
- graph mining
- graph clustering
- bipartite graph
- subgraph isomorphism
- graph properties
- noisy environments
- graph structures
- spanning tree
- structural pattern recognition
- graph transformation
- random graphs
- graph isomorphism
- adjacency matrix
- graph kernels
- graph classification
- minimum spanning tree
- series parallel
- graph search
- graph data
- undirected graph
- artificial neural networks
- evolving graphs
- active learning
- maximum clique
- graph patterns
- densely connected
- graphical models
- random walk
- sparse representation
- topological information
- query graph
- labeled data
- structured data
- bounded treewidth
- dynamic graph
- high dimensional
- pairwise
- label propagation
- training data