Node Classification 벤치마크
Node Classification on Wiki-CS
Accuracy
- 2023-08-17 — GraphSAGE: Accuracy 83.67
- 2023-12-28 — CGT: Accuracy 84.61
| Rank | Model | Accuracy | Paper | Code | Year |
|---|---|---|---|---|---|
| 1 | CGT | 84.61±0.53 | Mitigating Degree Biases in Message Passing Mechanism by Utilizing Community Structures | nslab-cuk/community-aware-graph-transformer | 2023 |
| 2 | GraphSAGE | 83.67 | Half-Hop: A graph upsampling approach for slowing down message passing | nerdslab/halfhop | 2023 |
| 3 | HH-GraphSAGE | 82.81 | Half-Hop: A graph upsampling approach for slowing down message passing | nerdslab/halfhop | 2023 |
| 4 | HH-GCN | 82.57 | Half-Hop: A graph upsampling approach for slowing down message passing | nerdslab/halfhop | 2023 |
| 5 | GCN | 81.93 | Half-Hop: A graph upsampling approach for slowing down message passing | nerdslab/halfhop | 2023 |
| 6 | ScaleNet | 79.3±0.6 | Scale Invariance of Graph Neural Networks | qin87/scalenet | 2024 |