Research map / Graph learning / Graph self-supervised learning
MV-FGAD: Towards Efficient and Effective Federated Graph Anomaly Detection via Multi-view Learning
Junyi Yan, KE LIANG, Hao Yu, Meng Liu, Hao Tan, Tianrui Liu, Jun-Jie Huang, Xinwang Liu
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 |
|---|---|---|---|
| use Mahalanobis distance instead of Euclidean distance for scoring (ablation) | MV-FGAD · Amazon | AUROC: 71.11 -> 78.33 | Improved |
| use Mahalanobis distance instead of cosine similarity for scoring (ablation) | MV-FGAD · Reddit | AUROC: 55.66 -> 59.27 | Improved |
| use Mahalanobis distance instead of cosine similarity (ablation) | MV-FGAD · Tolokers | AUPRC: 23.31 -> 23.56 | Improved |
| combine any two views of AAL, SAL, WADL (ablation) | MV-FGAD · Reddit, Tolokers, Amazon | AUROC | Improved |
| add all three views (AAL, SAL, WADL) (ablation) | MV-FGAD · Amazon | AUROC: 75.62 -> 78.33 | Improved |
| add weak anomaly-aware dynamic learning view (ablation) | MV-FGAD · multiple datasets | AUROC | Improved |
| use mean aggregation of multi-view scores (ablation) | MV-FGAD · multiple datasets | stability | Improved |
| use MV-FGAD instead of federated GAD baselines | MV-FGAD · Amazon-all | AUPRC: 20.61 -> 40.53 | Improved |
| use MV-FGAD instead of LG-FGAD | MV-FGAD · YelpChi | AUPRC: 33.36 -> 22.60 | Mixed |
| set alpha and beta to 0.5 (ablation) | MV-FGAD · multiple datasets | AUROC | Improved |
Closest papers by meaning
- Federated Graph Anomaly Detection Through Contrastive Learning with Global Negative Pairs (AAAI 2025 · similarity 0.91)
- ARC: A Generalist Graph Anomaly Detector with In-Context Learning (NeurIPS 2024 · similarity 0.88)
- Open-Set Graph Anomaly Detection via Normal Structure Regularisation (ICLR 2025 · similarity 0.88)
- Graph Anomaly Detection as Dynamical Transport: Training-Free Scoring via Empirical Bayes (NeurIPS 2026 · similarity 0.88)
- Generative Semi-supervised Graph Anomaly Detection (NeurIPS 2024 · similarity 0.87)
- UniGAD: Unifying Multi-level Graph Anomaly Detection (NeurIPS 2024 · similarity 0.86)
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.
