Personalized Federated Learning Unlearning on CIFAR-100 (alpha_dir=0.5)
100PPSRetrain
Evaluation results
| Method | Links | ||
|---|---|---|---|
| RetrainpFL Architecture=FedPer2026.06 | 100 | 100 | |
| RetrainpFL Architecture=FedRep2026.06 | 100 | 100 | |
| RetrainpFL Architecture=Ditto2026.06 | 100 | 100 | |
| RetrainpFL Architecture=FedBN2026.06 | 100 | 100 | |
| pFedULpFL Architecture=Ditto2026.06 | 97.2 | 95.6 | |
| pFedULpFL Architecture=FedRep2026.06 | 96.8 | 95 | |
| pFedULpFL Architecture=FedPer2026.06 | 96.3 | 94.4 | |
| pFedULpFL Architecture=FedBN2026.06 | 95.9 | 93.8 | |
| Fine-tunepFL Architecture=Ditto2026.06 | 93.9 | 90.1 | |
| Fine-tunepFL Architecture=FedRep2026.06 | 93.2 | 89.3 | |
| Fine-tunepFL Architecture=FedPer2026.06 | 92.8 | 88.6 | |
| Fine-tunepFL Architecture=FedBN2026.06 | 92.1 | 87.8 | |
| FedEraser-adapt.pFL Architecture=Ditto2026.06 | 85.8 | 82.8 | |
| FedEraser-adapt.pFL Architecture=FedRep2026.06 | 84.9 | 81.6 | |
| FedEraser-adapt.pFL Architecture=FedPer2026.06 | 84.1 | 80.7 | |
| FedEraser-adapt.pFL Architecture=FedBN2026.06 | 83.2 | 79.5 | |
| GA-NaivepFL Architecture=Ditto2026.06 | 81.2 | 76.8 | |
| GA-NaivepFL Architecture=FedRep2026.06 | 80.1 | 75.2 | |
| GA-NaivepFL Architecture=FedPer2026.06 | 79.3 | 74.1 | |
| GA-NaivepFL Architecture=FedBN2026.06 | 78.4 | 72.9 |