![[2602.20019] Learning Discriminative and Generalizable Anomaly Detector for Dynamic Graph with Limited Supervision](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.20019] Learning Discriminative and Generalizable Anomaly Detector for Dynamic Graph with Limited Supervision
Summary
This article presents a novel framework for dynamic graph anomaly detection that effectively utilizes limited labeled anomalies while maintaining generalization to unseen anomalies.
Why It Matters
The study addresses a significant challenge in anomaly detection within dynamic graphs, which is crucial for various applications such as fraud detection and network security. By proposing a model-agnostic approach that balances the use of limited labeled data and generalization, this research contributes to advancing machine learning techniques in real-world scenarios.
Key Takeaways
- Introduces a framework for dynamic graph anomaly detection using limited supervision.
- Utilizes residual representation encoding to capture deviations in interactions.
- Employs a restriction loss to maintain consistent scales for normal representations.
- Implements a bi-boundary optimization strategy for robust anomaly detection.
- Demonstrates superior performance across various evaluation settings.
Computer Science > Machine Learning arXiv:2602.20019 (cs) [Submitted on 23 Feb 2026] Title:Learning Discriminative and Generalizable Anomaly Detector for Dynamic Graph with Limited Supervision Authors:Yuxing Tian, Yiyan Qi, Fengran Mo, Weixu Zhang, Jian Guo, Jian-Yun Nie View a PDF of the paper titled Learning Discriminative and Generalizable Anomaly Detector for Dynamic Graph with Limited Supervision, by Yuxing Tian and 5 other authors View PDF HTML (experimental) Abstract:Dynamic graph anomaly detection (DGAD) is critical for many real-world applications but remains challenging due to the scarcity of labeled anomalies. Existing methods are either unsupervised or semi-supervised: unsupervised methods avoid the need for labeled anomalies but often produce ambiguous boundary, whereas semi-supervised methods can overfit to the limited labeled anomalies and generalize poorly to unseen anomalies. To address this gap, we consider a largely underexplored problem in DGAD: learning a discriminative boundary from normal/unlabeled data, while leveraging limited labeled anomalies \textbf{when available} without sacrificing generalization to unseen anomalies. To this end, we propose an effective, generalizable, and model-agnostic framework with three main components: (i) residual representation encoding that capture deviations between current interactions and their historical context, providing anomaly-relevant signals; (ii) a restriction loss that constrain the normal representations...
