What type of study is this?

September 10, 2025Applied SciencesOpen Access

Enhanced Graph Autoencoder for Graph Anomaly Detection Using Subgraph Information

View Full Paper

Ask AI

Bookmark

Authors

Observational analysis leverages subgraph information to improve anomaly detection in attributed networks, suggesting superior model effectiveness.

The proposed framework improves graph anomaly detection by utilizing subgraph information for better representation learning.
Enhanced detection performance is achieved through the incorporation of a graph structure learning-based decoder.
Model design emphasizes unsupervised learning, addressing the scarcity of labeled anomalies in real-world datasets.
Comprehensive experiments on six real-world datasets highlight the framework's effectiveness over existing methods.

Zhang et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

Observational analysis leverages subgraph information to improve anomaly detection in attributed networks, suggesting superior model effectiveness.