Generalization bound certification on CIFAR10 (test)
3.8Generalization BoundNorm-based (S5)
Evaluation results
| Method | Links | |
|---|---|---|
| Norm-based (S5)Bound (Theorem)=S5, Approach type=Baselines, Target Model=ResNet182026.02 | 3.8 | |
| Random Coreset (S8)Bound (Theorem)=S8, Approach type=Our approach, Target Model=ResNet182026.02 | 17.36 | |
| PAC-Bayes (S15)Bound (Theorem)=S15, Approach type=Our approach, Target Model=ResNet182026.02 | 28.13 | |
| Model compression (S11)Bound (Theorem)=S11, Approach type=Our approach, Target Model=ResNet182026.02 | 35.06 | |
| Pick-To-Learn (S10)Bound (Theorem)=S10, Approach type=Our approach, Target Model=ResNet182026.02 | 57.13 | |
| Best Coreset (S9)Bound (Theorem)=S9, Approach type=Our approach, Target Model=ResNet182026.02 | 81.72 | |
| Partition-based (S4)Bound (Theorem)=S4, Approach type=Baselines, Target Model=ResNet182026.02 | 90.58 |