CaT-GNN: Enhancing Credit Card Fraud Detection via Causal Temporal Graph Neural Networks
By Yifan Duan, Guibin Zhang, Shilong Wang, Xiaojiang Peng, Wang Ziqi, Junyuan Mao, Hao Wu, Xinke Jiang, Kun Wang
Rating
1564
Battle Count: 31
Relevance
7/10
While focused on fraud detection, the causal temporal graph approach could be adapted for analyzing trading networks and detecting market anomalies.
Implementation Complexity
8/10
Requires understanding of graph neural networks, causal inference, and temporal data processing. Implementation of custom layers and causal intervention mechanisms adds complexity.
Reproducibility
4/5
The paper provides detailed experimental setup, hyperparameters, and dataset information. Code availability is not mentioned explicitly.
About this paper
Methodology: Causal Temporal Graph Neural Network (CaT-GNN). Problem types: Classification, Anomaly Detection, Graph Learning.
The interactive Everscope explorer (charts, battles, favorites) loads below.