CSGNN: Improving Graph Neural Networks with Contrastive Semi-supervised Learning.
Yumeng SongYu GuXiaohua LiChuanwen LiGe YuPublished in: DASFAA (1) (2022)
Keyphrases
- semi supervised learning
- neural network
- graph based semi supervised learning
- label propagation
- graph construction
- semi supervised classification
- semi supervised
- unlabeled data
- labeled data
- label propagation algorithm
- supervised learning
- pattern recognition
- unsupervised learning
- machine learning
- co training
- manifold regularization
- active learning
- training data
- semi supervised learning algorithms
- weighted graph
- graph laplacian
- domain adaptation
- improve the classification accuracy
- transfer learning
- graph structure
- labeled and unlabeled data
- regularization framework
- random walk
- multi view
- learning models
- k nearest neighbor
- text classification
- prior knowledge
- semi supervised learning methods
- reinforcement learning