Fraud Detection is defined as the identification of suspicious transfers, orders and other illicit operations directed at the company.

There are two types of Fraud Detection models, application fraud models and behavioral (transactional) fraud models. Application models are used to identify "suspect customers" in an early stage of signing the contract. These models rely on data obtained from the application form. On the other hand, behavioral models are constructed on the basis of the data gathered during the period when the customer was active, such as transaction data, the information about used services or other kinds of activities.

StatConsulting offers various Fraud Scoring models and Fraud Scoring engines. Those solutions provide the means to:

• identify fraudulent operations in real time,
• determine the factors influencing the occurrence of a fraudulent operation,
• characterize the profile of a customer likely to perform a fraud,
• decrease losses due to fraudulent operations.

See also:

• Marketing Support and Analytical CRM