Node Classification on Brazil
80.8AccuracyMC-GPB (+)
Evaluation Results
| Method | Links | |
|---|---|---|
| MC-GPB (+)2026.06 | 80.8 | |
| UGTpre-training=true2023.08 | 80 | |
| UGTpre-training=false2023.08 | 76.92 | |
| No defense2026.06 | 76.9 | |
| GATv22023.08 | 69.23 | |
| DeeperGCN2023.08 | 69.23 | |
| SAGE2023.08 | 66.15 | |
| SAT2023.08 | 66.15 | |
| Random noise2026.06 | 63.4 | |
| GT2023.08 | 63.07 | |
| SAN2023.08 | 61.53 | |
| DEMO-Netvariant=hash2019.06 | 61.4 | |
| GAT2023.08 | 58.46 | |
| WRGAT2023.08 | 55.38 | |
| DEMO-Netvariant=weight2019.06 | 54.3 | |
| GPS2023.08 | 52.31 | |
| GCN cheby2019.06 | 51.6 | |
| GIN2023.08 | 50.76 | |
| ANS-GT2023.08 | 46.8 | |
| Union2019.06 | 46.6 | |
| Intersection2019.06 | 45.9 | |
| GCN2019.06 | 43.2 | |
| Differential privacy2026.06 | 42.3 | |
| Graphormer2023.08 | 42 | |
| GCN2023.08 | 41.54 | |
| GraphSAGE2019.06 | 40.4 | |
| GAT2019.06 | 38.2 | |
| MC-GPB (+)Defense Method=MC-GPB (+), Backbone Architecture=GPR-GNN2026.06 | 26.9 | |
| GCNII2023.08 | 24.35 |