Confidence May Cheat: Self-Training on Graph Neural Networks under Distribution Shift.
Hongrui LiuBinbin HuXiao WangChuan ShiZhiqiang ZhangJun ZhouPublished in: CoRR (2022)
Keyphrases
- neural network
- semi supervised classification
- pattern recognition
- artificial neural networks
- connected components
- graph matching
- fuzzy logic
- semi supervised
- multi class
- back propagation
- spatial distribution
- bipartite graph
- graph theory
- graph structure
- spanning tree
- training process
- confidence intervals
- semi supervised learning
- graph partitioning
- probability distribution
- input pattern
- graph databases
- feed forward
- random graphs
- degree distribution
- directed acyclic graph
- multilayer perceptron
- data distribution
- random walk
- graphical models
- association rules
- image segmentation
- learning algorithm
- genetic algorithm