Research map / Time series / Anomaly detection
Deep Variational Graph Convolutional Recurrent Network for Multivariate Time Series Anomaly Detection
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 prediction (forecasting) loss alongside reconstruction loss (VGCRN vs VGCRN-rec) (ablation) | VGCRN · DND | F1-score: 0.820 -> 0.826 | Helped |
| add prediction loss alongside reconstruction loss (VGCRN vs VGCRN-rec) (ablation) | VGCRN · SMD | F1-score: 0.893 -> 0.901 | Helped |
| stack to 3 layers (DVGCRN-layer3 vs single-layer VGCRN) | DVGCRN · DND | F1-score: 0.826 -> 0.834 | Helped |
| use modified multi-layer anomaly score (Eq.14, '-M') (ablation) | DVGCRN-layer3 · DND | F1-score: 0.834 -> 0.849 | Helped |
| use modified multi-layer anomaly score (Eq.14, '-M') (ablation) | DVGCRN-layer3 · SMD | F1-score: 0.912 -> 0.915 | Helped |
| use modified multi-layer anomaly score (Eq.14, '-M') (ablation) | DVGCRN-layer3 · MSL | F1-score: 0.908 -> 0.914 | Helped |
| use modified multi-layer anomaly score (Eq.14, '-M') (ablation) | DVGCRN-layer3 · SMAP | F1-score: 0.912 -> 0.914 | Helped |
| add channel graph (embedding-based adjacency) and recurrence/hierarchy in ablation study (ablation) | DVGCRN · DND and SMD | F1-score: qualitative bar chart only | Helped |
| increase network depth with fixed first-layer width (ablation) | DVGCRN · DND | F1-score: qualitative trend | Helped |
| increase hidden width beyond a moderate size at fixed depth (ablation) | DVGCRN · DND | F1-score: first increases then decreases | Mixed |
| increase number of layers beyond the best depth (ablation) | DVGCRN · DND and SMD | F1-score: not always improving with more layers | Mixed |
| weight reconstruction loss higher (gamma balance) (ablation) | DVGCRN · DND | F1-score: qualitative | Helped |
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 0.87)
- GCAD: Anomaly Detection in Multivariate Time Series from the Perspective of Granger Causality (AAAI 2025 · similarity 0.87)
- VACE: Learning Geometrically Structured Representations for Time Series Anomaly Detection (NeurIPS 2026 · similarity 0.86)
- MPTSNet: Integrating Multiscale Periodic Local Patterns and Global Dependencies for Multivariate Time Series Classification (AAAI 2025 · similarity 0.85)
- Structured Temporal Causality for Interpretable Multivariate Time Series Anomaly Detection (NeurIPS 2025 · 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.