Machine Unlearning on Tiny-Imagenet Random Forget 10%, γ=1 (test)
96.33FAl1-sparse
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| l1-sparseBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 96.33 | 95.47 | 55.19 | 3.67 | 38.07 | |
| IUBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 96.29 | 95.46 | 55.71 | 3.71 | 38.18 | |
| FTBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 86.97 | 99.98 | 53.95 | 13.03 | 31.95 | |
| GABackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 77.98 | 93.73 | 48.67 | 22.02 | 27.77 | |
| BEBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 74.92 | 92.61 | 45.89 | 25.08 | 27.22 | |
| BSBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 72.6 | 92.42 | 46.25 | 27.4 | 26.02 | |
| FalWBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 25.7 | 99.98 | 49.11 | 73.3 | 0.35 | |
| RetrainBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 25.63 | 99.98 | 48.83 | 74.36 | 0 | |
| SalUnBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 22.68 | 99.18 | 46.11 | 71.02 | 2.45 | |
| SFRonBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 22.01 | 97.1 | 44.23 | 77.98 | 3.68 | |
| RLBackbone=Vgg-16, Random Forget Ratio=10%, γ (Imbalance Factor)=12026.01 | 18.5 | 99.83 | 48.51 | 81.5 | 3.69 |