Time-Series Anomaly Detection on PSM (F1, AUROC, AUPR, VUS-ROC, VUS-PR)
51.2F1 ScoreASTER
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| ASTERModel Category=LLM-based2026.04 | 51.2 | 69.7 | 50.1 | 63.1 | 45.2 | |
| TranADModel Category=Deep Learning-based2026.04 | 40.3 | 63.1 | 38.8 | 56.6 | 43.1 | |
| LLMMixerModel Category=LLM-based2026.04 | 18.6 | 59.3 | 38.5 | 57.9 | 38.3 | |
| TimerModel Category=Pre-Trained Models2026.04 | 14.3 | 55.6 | 35.9 | 54.1 | 35 | |
| ChronosModel Category=Pre-Trained Models2026.04 | 13.1 | 56.5 | 37 | 54.7 | 36.2 | |
| MomentModel Category=Pre-Trained Models2026.04 | 11.7 | 54.5 | 33.2 | 54 | 32.9 | |
| CALFModel Category=LLM-based2026.04 | 10.2 | 58.9 | 37 | 58.7 | 37.4 | |
| UniTimeModel Category=LLM-based2026.04 | 9.7 | 57.9 | 37.7 | 57.9 | 37.7 | |
| GPT4TSModel Category=LLM-based2026.04 | 8.9 | 58 | 37.6 | 57.6 | 37.4 | |
| TimesNetModel Category=Deep Learning-based2026.04 | 8.8 | 59.2 | 39.1 | 59.3 | 39.5 | |
| TimesFMModel Category=Pre-Trained Models2026.04 | 8.2 | — | — | — | — | |
| DAGMMModel Category=Deep Learning-based2026.04 | 0.7 | 63.7 | 41.6 | 60.8 | 40.4 |