Supervised machine learning algorithms predict outcomes and deliver risk scores in real time, pulling from billions of global transactions and 80M+ detailed fraud and abuse reports.