Research map / Time series / Anomaly detection
Precision and Recall for Time Series
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 |
|---|---|---|---|
| Use front positional bias for recall instead of back bias (ablation) | Range-based F1_T · Front-Predicted | F1_T score: 0.11 -> 0.42 | Helped |
| Use back positional bias for recall instead of front bias (ablation) | Range-based F1_T · Back-Predicted | F1_T score: 0.11 -> 0.42 | Helped |
| Use range-based recall with positional bias rather than classical point-based recall | LSTM-AD · Time-Guided | recall | Mixed |
| Use range-based precision rather than classical point-based precision | LSTM-AD · NYC-Taxi | precision | Helped |
| Use range-based precision rather than classical point-based precision | LSTM-AD · Most evaluated datasets | precision | Hurt |
| Use the range-based model with front-biased recall and flat-biased precision instead of Numenta scoring | Anomaly scoring · Front-Predicted and Back-Predicted | sensitivity to prediction position: Numenta_Standard: 0.67 vs 0.63; F1_T_Front_Flat: 0.42 vs 0.11; F1_T_Back_Flat: 0.11 vs 0.42 | Helped |
| Apply cardinality penalty to fragmented predictions | Range-based precision and recall · ECG | F1 score | Hurt |
| Evaluate detectors with range-based scoring rather than classical and Numenta scoring | LSTM-AD, Greenhouse, and Luminol · ECG | detector ranking | Mixed |
| Evaluate detectors with range-based scoring rather than classical and Numenta scoring | LSTM-AD, Greenhouse, and Luminol · NYC-Taxi | detector ranking | Mixed |
| Use the model's classical-equivalent settings for unit-size ranges (ablation) | Range-based precision and recall · Unit-size ranges | precision and recall | No clear effect |
Closest papers by meaning
- The Elephant in the Room: Towards A Reliable Time-Series Anomaly Detection Benchmark (NeurIPS 2024 · similarity 0.87)
- Position: Quo Vadis, Unsupervised Time Series Anomaly Detection? (ICML 2024 · similarity 0.86)
- Retrieval Over Training: Similarity search-based Model Selection for Time Series Anomaly Detection (NeurIPS 2026 · similarity 0.85)
- When Model Meets New Normals: Test-Time Adaptation for Unsupervised Time-Series Anomaly Detection (AAAI 2024 · similarity 0.85)
- Supervision Recovery for Time Series Anomaly Detection via Counterfactual Pairing (NeurIPS 2026 · similarity 0.84)
- Detect, Explain, Interpret: An End-to-End Benchmark for Time Series Anomaly Detection, Explainability and Interpretability. (NeurIPS 2026 · similarity 0.84)
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.