Federated Unlearning Accuracy on CIFAR-10
68.7Pretrain Accuracyf-FUM (KL-KL)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| f-FUM (KL-KL)N (Number of clients)=5, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 68.7 | 73 | |
| f-FUM (KL-JS)N (Number of clients)=5, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 68.7 | 76.62 | |
| f-FUM (KL-χ²)N (Number of clients)=5, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 68.7 | 72.12 | |
| NoTN (Number of clients)=5, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 68.7 | 72 | |
| HalimiN (Number of clients)=5, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 68.7 | 74 | |
| f-FUM (KL-KL)N (Number of clients)=10, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 67.02 | 70.66 | |
| f-FUM (KL-JS)N (Number of clients)=10, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 67.02 | 73.98 | |
| f-FUM (KL-χ²)N (Number of clients)=10, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 67.02 | 69.1 | |
| NoTN (Number of clients)=10, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 67.02 | 70 | |
| HalimiN (Number of clients)=10, Removed (%)=2%, Backbone=ResNet-18, Unlearning Scenario=Confuse2026.02 | 67.02 | 70 | |
| f-FUM (KL-KL)N=10, Removed (%)=102026.02 | 38.42 | 52.74 | |
| f-FUM (KL-JS)N=10, Removed (%)=102026.02 | 38.42 | 72.32 | |
| f-FUM (KL-chi2)N=10, Removed (%)=102026.02 | 38.42 | 52.34 | |
| NoTN=10, Removed (%)=102026.02 | 38.42 | 42 | |
| HalimiN=10, Removed (%)=102026.02 | 38.42 | 42 | |
| f-FUM (KL-KL)N=10, Removed (%)=62026.02 | 28.42 | 52.62 | |
| f-FUM (KL-JS)N=10, Removed (%)=62026.02 | 28.42 | 73.74 | |
| f-FUM (KL-chi2)N=10, Removed (%)=62026.02 | 28.42 | 44.78 | |
| NoTN=10, Removed (%)=62026.02 | 28.42 | 43 | |
| HalimiN=10, Removed (%)=62026.02 | 28.42 | 43 |