DGFraud

DGFraud is a Graph Neural Network (GNN) toolbox designed for detecting fraud in various systems by integrating and comparing state-of-the-art GNN-based models. Its primary use case lies in enhancing the efficacy of fraud detection mechanisms through advanced graph-based methodologies. Notable features include a modular architecture for implementing new models, comprehensive documentation on existing algorithms, and support for TensorFlow 2.0, allowing seamless integration into existing projects.