Mastercard Incorporated
SYSTEMS AND METHODS FOR MODELING AND CLASSIFICATION OF FRAUDULENT TRANSACTIONS

Abstract:

Described herein are systems and methods for classifying incoming payment transactions. A fraud classification computing system includes a historical transaction database for storing a plurality of transaction records associated with a respective plurality of historical transactions. The fraud classification computing system receives a current transaction request message associated with a current payment transaction. The fraud classification computing system applies a multi-class fraud prediction model to the current transaction request message to generate scores indicating a relative likelihood that the current payment transaction is each of a plurality of fraudulent transaction types. The fraud classification computing system identifies a most likely transaction classification identifier and generates a transaction classification message for the current payment transaction.