Versions#

V2.0.0#

Features#

  • Edge Prediction

  • Builds a GNN Model

  • Train an XGBoost model to predict fraud scores

  • Uses Wholegraph and cuGraph to encapsulate graph storage and complexity efficiently

  • Support to use XGBoost modeling optionally without GNN embeddings

V1.0.0#

Features#

  • Node Prediction

  • Builds a GNN Model

  • Train an XGBoost model to predict fraud scores

  • Uses Wholegraph and cuGraph to encapsulate graph storage and complexity efficiently

  • Support to use XGBoost modeling optionally without GNN embeddings