Malicious Traffic Classification on Malicious Traffic Dataset (Internal Seen)
99.74Benign SpecificityMLP
Evaluation results
| Method | Links | |
|---|---|---|
| MLPArchitecture=MLP, Evaluation Protocol=Imbalanced test set (~4:1 Benign)2025.12 | 99.74 | |
| CNNArchitecture=CNN, Evaluation Protocol=Imbalanced test set (~4:1 Benign)2025.12 | 98.22 | |
| Latent Sculpting (Stage 1)Method Stage=Stage 1 (Encoder Only), Evaluation Protocol=Strictly balanced 1:1 test set2025.12 | 94.84 | |
| Latent Sculpting (Two-Stage)Sensitivity Threshold (P)=P99, Evaluation Protocol=Strictly balanced 1:1 test set2025.12 | 94 | |
| OCSVMArchitecture=OCSVM, Evaluation Protocol=Imbalanced test set (~4:1 Benign)2025.12 | 93.16 | |
| Latent Sculpting (Two-Stage)Sensitivity Threshold (P)=P97, Evaluation Protocol=Strictly balanced 1:1 test set2025.12 | 92 | |
| Latent Sculpting (Two-Stage)Sensitivity Threshold (P)=P95, Evaluation Protocol=Strictly balanced 1:1 test set2025.12 | 90 | |
| LOFArchitecture=LOF, Evaluation Protocol=Imbalanced test set (~4:1 Benign)2025.12 | 83.41 |