Research map / Time series / Anomaly detection
Detecting Multivariate Time Series Anomalies with Zero Known Label
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add dynamic graph structure learning (ablation) | MTGFlow · SWaT | AUROC(%): 78.3\textpm{0.9} -> 81.2\textpm{1.1} | Helped |
| add dynamic graph structure learning (ablation) | MTGFlow · WADI | AUROC(%): 89.7\textpm{0.5} -> 91.0\textpm{0.7} | Helped |
| add entity-aware normalizing flow (ablation) | MTGFlow · SWaT | AUROC(%): 78.3\textpm{0.9} -> 82.4\textpm{1.0} | Helped |
| add entity-aware normalizing flow (ablation) | MTGFlow · WADI | AUROC(%): 89.7\textpm{0.5} -> 91.3\textpm{0.4} | Helped |
| combine graph structure learning with entity-aware flow (ablation) | MTGFlow · SWaT | AUROC(%): 78.3\textpm{0.9} -> 84.8\textpm{1.5} | Helped |
| combine graph structure learning with entity-aware flow (ablation) | MTGFlow · WADI | AUROC(%): 89.7\textpm{0.5} -> 91.9\textpm{1.1} | Helped |
| draw independent mean vectors for each entity's target distribution instead of using a shared element value (ablation) | MTGFlow entity-aware flow · MTS anomaly detection | AUROC | Hurt |
| increase the number of flow blocks at small window sizes (ablation) | MTGFlow · SWaT and WADI, window sizes 40, 60, and 80 | AUROC(%): SWaT window 60: 84.8\textpm{1.5} -> 83.6\textpm{2.0} -> 83.1\textpm{0.9}; WADI window 40: 90.8\textpm{1.3} -> 91.7\textp | Mixed |
| increase the number of flow blocks at larger window sizes (ablation) | MTGFlow · SWaT and WADI, window sizes 100 and 120 | AUROC(%): SWaT window 120: 83.2\textpm{2.0} -> 83.4\textpm{2.3} -> 84.5\textpm{2.6}; WADI window 100: 89.6\textpm{1.1} -> 90.9\tex | Mixed |
Closest papers by meaning
- Graph Mixture of Experts and Memory-augmented Routers for Multivariate Time Series Anomaly Detection (AAAI 2025 · similarity 0.89)
- Multi-Scale Dynamic Graph Learning for Time Series Anomaly Detection (Student Abstract) (AAAI 2024 · similarity 0.88)
- Structured Temporal Causality for Interpretable Multivariate Time Series Anomaly Detection (NeurIPS 2025 · similarity 0.88)
- The Elephant in the Room: Towards A Reliable Time-Series Anomaly Detection Benchmark (NeurIPS 2024 · similarity 0.87)
- TAIL-MIL: Time-Aware and Instance-Learnable Multiple Instance Learning for Multivariate Time Series Anomaly Detection (AAAI 2025 · similarity 0.87)
- GCAD: Anomaly Detection in Multivariate Time Series from the Perspective of Granger Causality (AAAI 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.