Transaction sequence processing with embedded real-time decision feedback
Abstract:
A system and method for scoring an interaction using an analytical model and authorization decisions is disclosed. The method includes receiving from an access device an authorization request message for an interaction between a user and a resource provider. An analytical model comprising a neural network with at least one long short-term memory determines a score based on data in the authorization request message. The analytical model was formed using interaction data from prior authorization request messages, and authorization response messages from an authorizing computer. The authorization request message and the score is transmitted to the authorizing computer and an authorization response message, including an indication of whether the interaction was approved or declined, is received. Then the analytical model is updated based upon data in the authorization request message and the indication in the authorization response message to form an updated analytical model.
Information query
Patent Agency Ranking
0/0