Federated Node Classification on Amazon Electronics
67.96AccuracyFedEPD
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| FedEPDBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 67.96 | 63.99 | 65.09 | |
| GraphSMOTEBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 64.04 | 61.37 | 60.45 | |
| LWSBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 63.83 | 59.89 | 59.42 | |
| FedGTABackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 62.89 | 59.92 | 60.07 | |
| FedAvgBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 62.33 | 59.3 | 59.15 | |
| FedSprayBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 61.65 | 56.25 | 56.32 | |
| HieTailBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 61 | 56.48 | 56.2 | |
| CReFFBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 58.67 | 52.77 | 53.3 | |
| FedLCBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 58.37 | 52.39 | 52.57 | |
| FedGACBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 57.05 | 57.34 | 54.31 | |
| GraphFedMigBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 47.71 | 40.5 | 41.95 | |
| FedProtoBackbone=two-layer GCN, Hidden Dimension=64, Learning Rate=1e-2, Communication Rounds=200, Local Epochs=3, Number of Clients=102026.06 | 39.14 | 32.89 | 32.66 |