Abstract: Detecting anomalies on graph data has two types of methods. One is pattern mining that discovers strange structures globally such as quasi-cliques, bipartite cores, or dense blocks in the ...

Abstract: Anomaly detection significantly enhances the robustness of cloud systems. While neural network-based methods have recently demonstrated strong advantages, they encounter practical challenges ...