Temporal Graph Networks for Graph Anomaly Detection in Financial Networks
By Yejin Kim, Youngbin Lee, Minyoung Choe, Sungju Oh, Yongjae Lee
Rating
1420
Battle Count: 9
Relevance
8/10
Highly relevant for fraud detection and risk management in financial networks, which are crucial aspects of quantitative trading systems.
Implementation Complexity
7/10
Implementing TGNs and various graph embedding techniques requires advanced knowledge of graph neural networks and temporal data processing.
Reproducibility
3/5
The paper provides details on the dataset, model architectures, and hyperparameters, but code is not explicitly mentioned as being available.
About this paper
Methodology: Temporal Graph Networks. Problem types: Graph Anomaly Detection, Node Classification, Fraud Detection.
The interactive Everscope explorer (charts, battles, favorites) loads below.