BEHAVIOR BASED CREDIT CARD FRAUD DETECTION USING SUPPORT VECTOR MACHINES

• Main Authors: V. Dheepa, R. Dhanapal
• Format: Article
• Language: English
• Published: ICT Academy of Tamil Nadu 2012-07-01
• Series: ICTACT Journal on Soft Computing
• Subjects: Data Mining; Classification; Fraud Detection; Support Vector Machine; E-commerce
• Online Access: http://ictactjournals.in/paper/IJSCP7_391_397.pdf

Along with the great increase of internet and e-commerce, the use of credit card is an unavoidable one. Due to the increase of credit card usage, the frauds associated with this have also increased. There are a lot of approaches used to detect the frauds. In this paper, behavior based classification approach using Support Vector Machines are employed and efficient feature extraction method also adopted. If any discrepancies occur in the behaviors transaction pattern then it is predicted as suspicious and taken for further consideration to find the frauds. Generally credit card fraud detection problem suffers from a large amount of data, which is rectified by the proposed method. Achieving finest accuracy, high fraud catching rate and low false alarms are the main tasks of this approach.