Improved Sheep Flock Heredity Algorithm Based Prevention of Credit Card Fraud Detection for Online and Offline Transaction

摘要

One of the Intrusion Detection System is credit card fraud detection in data mining. The existing approaches validate the fraud occurrence by computing a communal analysis suspicion score of the credit applications. The key challenge of this paper is, to improve the efficiency of the credit card fraud detection method by verifying and validating the optimized parameters such as single and multiple attributes. The attributes of every application [offline/online] are verified using a newly developed procedure is MLMA-[Multi-Level-Multi-Agent] and it is verified all the attribute values are best one or not. For optimizing the attributes the ISFH-[Improved Sheep Flock Heredity] algorithm is used and those attributes are validated according to the time and response with optimal value. The experimental results of the proposed approach are compared with the existing approach results to compute the performance evaluation where the proposed approach experiments in DOTNET framework 2012 software.