Anomaly Detection on FSB sequence type
0.88Performance (Sine Sequence)tcNF-base
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| tcNF-baseCategory=deep learning, Approach=distribution2026.03 | 0.88 | 0.72 | 0.97 | 0.95 | 0.94 | 0.85 | |
| tcNF-mlpCategory=deep learning, Approach=distribution2026.03 | 0.85 | 0.66 | 0.9 | 0.95 | 0.81 | 0.84 | |
| tcNF-cnnCategory=deep learning, Approach=distribution2026.03 | 0.85 | 0.67 | 0.92 | 0.96 | 0.92 | 0.85 | |
| PCACategory=classic machine learning, Approach=reconstruction2026.03 | 0.83 | 0.61 | 0.89 | 0.98 | 0.42 | 0.7 | |
| GDNCategory=deep learning, Approach=forecasting2026.03 | 0.83 | 0.56 | 0.81 | 0.82 | 0.85 | 0.76 | |
| tcNF-statelessCategory=deep learning, Approach=distribution2026.03 | 0.82 | 0.65 | 0.91 | 0.96 | 0.87 | 0.86 | |
| RealNVPCategory=deep learning, Approach=distribution2026.03 | 0.8 | 0.64 | 0.88 | 0.97 | 0.75 | 0.76 | |
| KNNCategory=classic machine learning, Approach=distance2026.03 | 0.78 | 0.62 | 0.14 | 0.96 | 0.76 | 0.63 | |
| CBLOFCategory=classic machine learning, Approach=distance2026.03 | 0.77 | 0.69 | 0.52 | 0.85 | 0.78 | 0.71 | |
| TorskCategory=deep learning, Approach=forecasting2026.03 | 0.75 | 0.74 | 0.51 | — | 0.85 | — | |
| iForestCategory=outlier detection, Approach=trees2026.03 | 0.75 | 0.56 | 0.75 | 0.91 | 0.7 | 0.67 | |
| COFCategory=classic machine learning, Approach=distance2026.03 | 0.74 | 0.77 | 0.6 | 0.86 | 0.81 | 0.69 | |
| IF-LOFCategory=outlier detection, Approach=trees2026.03 | 0.74 | 0.61 | 0.75 | 0.94 | 0.79 | 0.6 | |
| DAMPCategory=outlier detection, Approach=distance2026.03 | 0.69 | 0.7 | 0.87 | 0.93 | 0.57 | 0.64 | |
| HBOSCategory=outlier detection, Approach=distance2026.03 | 0.64 | 0.64 | 0.47 | 0.63 | 0.71 | 0.69 | |
| PCCCategory=classic machine learning, Approach=reconstruction2026.03 | 0.55 | 0.53 | 0.7 | 0.62 | 0.54 | 0.66 |