Welcome to Graphomaly’s documentation!

Graphomaly is a package for anomaly detection in graphs, especially in those modeling bank transactions. There are many possible applications, among which detecting money laundering schemes has a main position.

Contents:

• Data types
• Feature extraction
  • Graph to features
  • Transaction list to features
  • Transaction list to graph
• Graphomaly Preprocessing
  • Egonet
  • Random walk
  • Spectrum
  • Graph to features
  • Transactions to graph
  • Time-related features
  • Group statistical features
  • Historical features
• Graphomaly Models
  • Autoencoder
  • Dictionary Learning
  • Isolation Forest
  • Variational Autoencoder
• Graphomaly Estimator
  • Estimator
  • Grid Search
  • Voting Classifier

Indices and tables

• Index

• Module Index

• Search Page