Shift-Robust GNNs: Overcoming the Limitations of Localized Graph Training data.
Qi ZhuNatalia PonomarevaJiawei HanBryan PerozziPublished in: CoRR (2021)
Keyphrases
- training data
- graph representation
- data sets
- graph theoretic
- decision trees
- training set
- bipartite graph
- classification accuracy
- graph model
- test data
- test set
- manually labeled training data
- training instances
- graph partitioning
- graph structure
- noisy data
- directed graph
- computationally efficient
- random walk
- supervised learning
- domain knowledge
- high dimensional
- learning algorithm