Node Classification 벤치마크
Node Classification on AMZ Computers
Accuracy
- 2020-07-18 — DAGNN (Ours): Accuracy 84.5
- 2021-03-04 — CPF-ind-GAT: Accuracy 85.5
- 2023-06-04 — NCGCN: Accuracy 90.81
| Rank | Model | Accuracy | Paper | Code | Year |
|---|---|---|---|---|---|
| 1 | NCGCN | 90.81 ± 0.46 | Clarify Confused Nodes via Separated Learning | GISec-Team/NCGNN | 2023 |
| 2 | NCSAGE | 90.43 ± 0.72 | Clarify Confused Nodes via Separated Learning | GISec-Team/NCGNN | 2023 |
| 3 | CPF-ind-GAT | 85.5% | Extract the Knowledge of Graph Neural Networks and Go Beyond it: An Effective Knowledge Distillation Framework | BUPT-GAMMA/CPF | 2021 |
| 4 | DAGNN (Ours) | 84.5 ± 1.2 | Towards Deeper Graph Neural Networks | dmlc/dgl · mengliu1998/DeeperGNN · divelab/DeeperGNN | 2020 |
| 5 | GLNN | 83.03± 1.87% | Graph-less Neural Networks: Teaching Old MLPs New Tricks via Distillation | snap-research/graphless-neural-networks | 2021 |