Adversarial Robustness on CIC-IoT 2023 (test)
99.55Classification AccuracySHAP-based Model
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SHAP-based ModelAttacks=FGSM2025.11 | 99.55 | 99.48 | 0.52 | |
| SHAP-based ModelAttacks=PGD2025.11 | 99.55 | 100 | 0 | |
| SHAP-based ModelAttacks=DeepFool2025.11 | 99.55 | 99.04 | 0.96 | |
| Adversarially Trained ModelAttacks=FGSM2025.11 | 97.35 | 89.77 | 10.23 | |
| Adversarially Trained ModelAttacks=PGD2025.11 | 97.35 | 92.1 | 7.9 | |
| Adversarially Trained ModelAttacks=DeepFool2025.11 | 97.35 | 66.49 | 33.51 |