Node Classification on Telegram
96.8AccuracyScaleNet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ScaleNetType=Ours, 10-fold cross validation=true, loop_α, β, γ=1_0.5,-1,-12024.11 | 96.8 | — | |
| 1iGType=Symmetric (Ours), 10-fold cross validation=true2024.11 | 95.8 | — | |
| 1iGi2Type=Symmetric (Ours), 10-fold cross validation=true2024.11 | 93 | — | |
| 1iGu2Type=Symmetric (Ours), 10-fold cross validation=true2024.11 | 92.6 | — | |
| Dir-GNNType=BiDirection, 10-fold cross validation=true2024.11 | 90.2 | — | |
| DSHN2025.10 | 88.65 | — | |
| MagNetType=Hermitian, 10-fold cross validation=true2024.11 | 87.6 | — | |
| SymType=Symmetric, 10-fold cross validation=true2024.11 | 87.2 | — | |
| SigMaNetType=Hermitian, 10-fold cross validation=true2024.11 | 86.9 | — | |
| GCNType=Base models, 10-fold cross validation=true2024.11 | 86 | — | |
| QuaNetType=Hermitian, 10-fold cross validation=true2024.11 | 85.6 | — | |
| 1ymType=Symmetric (Ours), 10-fold cross validation=true2024.11 | 84 | — | |
| ChebNetType=Base models, 10-fold cross validation=true2024.11 | 83 | — | |
| DiGType=Symmetric, 10-fold cross validation=true2024.11 | 82 | — | |
| DSHNLight2025.10 | 81.15 | — | |
| DHGNNwith embeddings=true2025.10 | 80.58 | — | |
| DHGNN2025.10 | 79.62 | — | |
| GeDi-HNN2025.10 | 77.12 | — | |
| SAGEType=Base models, 10-fold cross validation=true2024.11 | 74 | — | |
| SheafHyperGNN2025.10 | 70 | — | |
| APPNPType=Base models, 10-fold cross validation=true2024.11 | 67.3 | — | |
| AllSetTransformer2025.10 | 66.92 | — | |
| DiGibType=Symmetric, 10-fold cross validation=true2024.11 | 64.1 | — | |
| ED-HNN2025.10 | 60.38 | — | |
| HGNN2025.10 | 51.73 | — | |
| UniGCNII2025.10 | 51.73 | — | |
| AllDeepSets2025.10 | 49.19 | — | |
| PhenomNN2025.10 | 47.69 | — | |
| LEGCN2025.10 | 45.38 | — | |
| HyperND2025.10 | 44.62 | — | |
| HNHN2025.10 | 38.22 | — | |
| MLPType=Base models, 10-fold cross validation=true2024.11 | 32.8 | — |