Table 5 Comparison of related one-class SVM anomaly detection results on 2 IEC 104 protocol datasets
From: Improving anomaly detection in SCADA network communication with attribute extension
FNR | FPR | MCC | AUC Score | ||
|---|---|---|---|---|---|
Dataset Egger et al. (2020) | Unsupervised learning\(^d\) Anwar et al. (2021) | 0.98 | 0.03 | \(-0.01\) | 0.49 |
Unsupervised learning\(^t\) Anwar et al. (2021) | 0.69 | 0.01 | 0.30 | 0.64 | |
Dataset Maynard et al. (2018) | Unsupervised learning\(^d\) (baseline) | 0.49 | 0.06 | 0.34 | 0.72 |
Unsupervised learning\(^d\) (endline) | 0.00 | 0.02 | 0.80 | 0.98 |