Multi-class Classification on UNSW-NB15 (test)
95.12F1 ScoreGTCN-G
Evaluation results
| Method | Links | |
|---|---|---|
| GTCN-GAttention mechanism=6-head, Dropout rate=0.52025.10 | 95.12 | |
| GATMode=Standard2025.10 | 91.78 | |
| E-GraphSAGE-MImplementation=mini-batching, Number of Layers=2, Aggregator=mean, Activation=ReLU, Sampling strategy=2-hop neighborhood, Sample size=8 per hop2025.10 | 89.34 | |
| E-GraphSAGE2025.10 | 87.56 |