A naukri.com initiative
Arxiv
2d
140
Image Credit: Arxiv
How to Use Graph Data in the Wild to Help Graph Anomaly Detection?
- Graph anomaly detection faces challenges like label scarcity and varying anomaly types, making supervised methods unreliable.
- Researchers propose Wild-GAD, a framework that utilizes external graph data to aid anomaly detection tasks.
- Wild-GAD is based on UniWildGraph database, containing diverse graph data to enhance anomaly detection accuracy.
- Extensive experiments show Wild-GAD outperforms baseline methods with an average 18% AUCROC and 32% AUCPR improvement.
Read Full Article
8 Likes
For uninterrupted reading, download the app