Multivariate Time-series Anomaly Detection on SMD 1-8
87.2AUROCCANDI
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CANDIalpha=0.5%, Backbone=MLP-based autoencoder2026.04 | 87.2 | 43.2 | 39.3 | |
| CANDIalpha=1.0%, Backbone=MLP-based autoencoder2026.04 | 87.2 | 43.4 | 40.9 | |
| CANDIalpha=5.0%, Backbone=MLP-based autoencoder2026.04 | 86.7 | 42.3 | 21.3 | |
| M2N2alpha=0.5%, Backbone=MLP-based autoencoder2026.04 | 83.7 | 40.7 | 40.6 | |
| M2N2alpha=1.0%, Backbone=MLP-based autoencoder2026.04 | 80.5 | 37.6 | 38.9 | |
| M2N2alpha=5.0%, Backbone=MLP-based autoencoder2026.04 | 77.2 | 35.4 | 11.5 | |
| Pretrainedalpha=0.5%, Backbone=MLP-based autoencoder2026.04 | 71.9 | 33.2 | 37.7 | |
| Pretrainedalpha=1.0%, Backbone=MLP-based autoencoder2026.04 | 71.9 | 33.2 | 36.2 | |
| Pretrainedalpha=5.0%, Backbone=MLP-based autoencoder2026.04 | 71.9 | 33.2 | 9.2 |