Machine Unlearning on CIFAR-100 superclass-wise
99.89Accuracy (Df)Original
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| OriginalDr=✓, Df=✓2025.08 | 99.89 | 99.99 | 76.63 | 1 | — | |
| RetrainDr=✓, Df=✗2025.08 | 0 | 99.53 | 74.05 | 100 | 0 | |
| IU*Dr=✓, Df=✓, Run time (s)=19±22025.08 | 0 | 69.29 | 61.85 | 100 | 4.82 | |
| GADr=✗, Df=✓, Run time (s)=9±02025.08 | 0 | 68.29 | 40.89 | 100 | 5.22 | |
| RLDr=✗, Df=✓, Run time (s)=22±32025.08 | 0 | 73.87 | 42.8 | 100 | 7.73 | |
| SSDDr=✓, Df=✓, Run time (s)=10±02025.08 | 0 | 77.37 | 46.85 | 100 | 3.15 | |
| SCAR*Dr=✗, Df=✓, Run time (s)=68±02025.08 | 0 | 77.79 | 60 | 100 | 7.13 | |
| NPODr=✗, Df=✓, Run time (s)=25±32025.08 | 0 | 90.47 | 55.8 | 100 | 5.35 | |
| IMUDr=✗, Df=✓, Run time (s)=14±02025.08 | 0 | 98.06 | 63.75 | 100 | 2.09 |