Research map / Label-efficient and robust learning / Semi-supervised learning
Learning Discriminative and Generalizable Anomaly Detector for Dynamic Graph with Limited Supervision
Yuxing Tian, Yiyan Qi, Fengran Mo, Weixu Zhang, Jian Guo, Jian-Yun Nie
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 residual representation encoding (ablation) | SDGAD on TCL · Wikipedia | F1: 5.29 ± 5.69 -> 8.34 ± 2.72 | Improved |
| add residual representation encoding (ablation) | SDGAD on TCL · MOOC | AUROC: 69.14 ± 2.48 -> 72.87 ± 1.40 | Improved |
| add representation restriction loss L_RR (ablation) | SDGAD on TCL · Wikipedia | F1: 6.01 ± 0.72 -> 8.34 ± 2.72 | Improved |
| add representation restriction loss L_RR (ablation) | SDGAD on TCL · Wikipedia | AUROC: 79.44 ± 0.80 -> 80.36 ± 0.69 | No clear effect |
| use bi-boundary optimization instead of single boundary (ablation) | SDGAD on TCL · Wikipedia | F1: 7.41 ± 3.29 -> 8.34 ± 2.72 | Improved |
| use bi-boundary optimization instead of single boundary (ablation) | SDGAD on TCL · MOOC | F1: 6.42 ± 0.97 -> 7.15 ± 0.64 | Improved |
| apply SDGAD framework to TCL encoder | TCL · Wikipedia | F1: 2.41 ± 0.0 -> 8.34 ± 2.72 | Improved |
| apply SDGAD framework to DyGFormer encoder | DyGFormer · Wikipedia | AUROC: 85.58 -> 86.60 | Improved |
| apply SDGAD to DyGFormer | DyGFormer · Wikipedia | AP: 2.58 -> 3.71 | Hurt |
| set decision margin tau to 0.1 instead of 0.05 (ablation) | SDGAD on TCL · Wikipedia | F1: 5.06 -> 8.34 | Improved |
| set decision margin tau to 0.1 instead of 0.2 (ablation) | SDGAD on TCL · Wikipedia | AUROC: 77.26 -> 80.36 | Improved |
| lower gamma from 0.99 to 0.95 (ablation) | SDGAD on TCL · Wikipedia | F1: 8.34 -> 3.51 | Hurt |
Closest papers by meaning
- ARC: A Generalist Graph Anomaly Detector with In-Context Learning (NeurIPS 2024 · similarity 0.91)
- Graph Anomaly Detection as Dynamical Transport: Training-Free Scoring via Empirical Bayes (NeurIPS 2026 · similarity 0.91)
- Leveraging Diffusion Model as Pseudo-Anomalous Graph Generator for Graph-Level Anomaly Detection (ICML 2025 · similarity 0.91)
- A Generalizable Anomaly Detection Method in Dynamic Graphs (AAAI 2025 · similarity 0.90)
- UniGAD: Unifying Multi-level Graph Anomaly Detection (NeurIPS 2024 · similarity 0.89)
- DiffGAD: A Diffusion-based Unsupervised Graph Anomaly Detector (ICLR 2025 · similarity 0.89)
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.
