Transductive Node Classification on Cora Full 5 labels per class
0.508Mean AccuracyCOLES-S^2GC
Evaluation results
| Method | Links | ||
|---|---|---|---|
| COLES-S^2GCLearning Paradigm=Contrastive2022.01 | 0.508 | 1.4 | |
| SCELearning Paradigm=Contrastive2022.01 | 0.507 | 1.5 | |
| DGILearning Paradigm=Contrastive2022.01 | 0.505 | 1.4 | |
| GCNLearning Paradigm=Semi-supervised2022.01 | 0.493 | 1.8 | |
| COLES-GCN (+Aug)Learning Paradigm=Contrastive + Augmentation2022.01 | 0.48 | 1.2 | |
| COLES-GCN (Stiefel)Learning Paradigm=Contrastive2022.01 | 0.476 | 1.2 | |
| MixHopLearning Paradigm=Semi-supervised2022.01 | 0.475 | 1.5 | |
| COLES-GCNLearning Paradigm=Contrastive2022.01 | 0.473 | 1.5 | |
| SGCLearning Paradigm=Unsupervised2022.01 | 0.46 | 2.2 | |
| GATLearning Paradigm=Semi-supervised2022.01 | 0.439 | 1.5 | |
| PCA-S^2GCLearning Paradigm=Unsupervised2022.01 | 0.423 | 1.7 | |
| S^2GCLearning Paradigm=Unsupervised2022.01 | 0.418 | 1.7 | |
| DeepWalkLearning Paradigm=Contrastive2022.01 | 0.389 | 1.4 | |
| Graph2GaussLearning Paradigm=Contrastive2022.01 | 0.389 | 1.3 | |
| SAGE with SampledNCELearning Paradigm=Contrastive2022.01 | 0.35 | 1.4 | |
| GCN with SampledNCELearning Paradigm=Contrastive2022.01 | 0.327 | 1.9 | |
| RP-S^2GCLearning Paradigm=Unsupervised2022.01 | 0.315 | 1.4 |