Research map / Time series / Anomaly detection
Time Series Anomaly Detection with Multiresolution Ensemble Decoding
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 multiresolution ensemble decoding instead of the strongest baseline | RAE-ensemble · ECG, 2D-gesture, Power-demand, and Yahoo’s S5 | AUROC: 0.6884 -> 0.7127; 0.7788 -> 0.8551; 0.8570 -> 0.8736; 0.6400 -> 0.6473; 0.8035 -> 0.8828; 0.5213 -> 0.6399; 0.7808 -> 0. | Helped |
| add coarse-to-fine fusion of lower-resolution decoder information (ablation) | RAMED · ECG(A), ECG(B), ECG(C), ECG(D), ECG(E), ECG(F), 2D-gesture, Power-demand, and Yahoo’s S5 | AUROC: 0.6781 -> 0.7127; 0.8196 -> 0.8551; 0.7537 -> 0.8736; 0.5125 -> 0.6473; 0.8212 -> 0.8828; 0.5083 -> 0.6399; 0.7656 -> 0. | Helped |
| add the multiresolution shape-forcing loss (ablation) | RAMED · ECG(A), ECG(B), ECG(C), ECG(D), ECG(E), ECG(F), 2D-gesture, Power-demand, and Yahoo’s S5 | AUROC: 0.6916 -> 0.7127; 0.8542 -> 0.8551; 0.8450 -> 0.8736; 0.5609 -> 0.6473; 0.8455 -> 0.8828; 0.5598 -> 0.6399; 0.7716 -> 0. | Helped |
| increase the decoder count from 1 to 3 (ablation) | RAMED · ECG(A) and 2D-gesture | AUROC: 0.6781 -> 0.7172; 0.7656 -> 0.7839 | Helped |
| increase the decoder count from 3 to 4, creating a coarsest decoder of length 2 (ablation) | RAMED · ECG(A) and 2D-gesture | AUROC: 0.7172 -> 0.6743; 0.7839 -> 0.7779 | Hurt |
| set the fusion weight β to 0.1 rather than larger tested values (ablation) | RAMED · ECG(A) and 2D-gesture | AUROC | Helped |
| use a small, nonzero shape-loss tradeoff λ = 10−4 (ablation) | RAMED · ECG(A) and 2D-gesture | AUROC | Helped |
Closest papers by meaning
- Towards a General Time Series Anomaly Detector with Adaptive Bottlenecks and Dual Adversarial Decoders (ICLR 2025 · similarity 0.86)
- Structured Temporal Causality for Interpretable Multivariate Time Series Anomaly Detection (NeurIPS 2025 · similarity 0.86)
- The Elephant in the Room: Towards A Reliable Time-Series Anomaly Detection Benchmark (NeurIPS 2024 · similarity 0.86)
- COMET: Codebook-based Online-adaptive Multi-scale Embedding for Time-series Anomaly Detection (NeurIPS 2026 · similarity 0.86)
- SARAD: Spatial Association-Aware Anomaly Detection and Diagnosis for Multivariate Time Series (NeurIPS 2024 · similarity 0.85)
- Supervision Recovery for Time Series Anomaly Detection via Counterfactual Pairing (NeurIPS 2026 · similarity 0.85)
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.