The Study of Fraud Detection in Financial and Credit Institutions with Real Data
This paper presents a review of data mining techniques for the fraud detection. Development of information systems such as data due to it has become a source of important organizations. Method and techniques are required for efficient access to data, sharing the data, extracting information from data and using this information. In recent years, data mining technology is an important method that it has changed to extract concepts from the data set. Scientific data mining and business intelligence technology is as a valuable and some what hidden to provide large volumes of data. This research studies using service analyzes software annual transactions related to 20000 account number of financial institutions in the country.The main data mining techniques used for financial fraud detection (FFD) are logistic models, neural networks and decision trees, all of which provide primarysolutions to the problems inherent in the detection and classification of fraudulent data. The proposed method is clustering clients based on client type. An appropriate rule for each cluster is determined by the behavior of group members in case of deviation from specified behavior will be known among suspected cases. The rules of the C5 have been applied in decision tree algorithm. Model is able to extract about a lot of the rules related to client behavior.
