Fraud Detection in Banking Using KNN (K-Nearest Neighbor) Algorithm
Fraud Detection in Banking Using KNN (K-Nearest Neighbor) Algorithm
masoud khodabakhshi1 mehdi fartash2
1) Department of Computer Engineering, Islamic Azad University, Arak Branch, Arak, Iran
2) Department of Computer Engineering, Islamic Azad University, Arak Branch, Arak, Iran Email:
IOI Code
:
XYFE-BBDZF
Publication :
5th International Conference on Research in Science & Technology(5rstconf.com)
Abstract :
Financial abuses and fraud in transaction banking has been increased because of using modern banking system. These abuses lose significant financial resources and decrease trust of customers in use of modern banking system and reduce effectiveness of these systems in optimum capital management and financial transactions. Although the best way to reduce fraud is preventing fraud but the fraudsters achieve their goals in some ways. So we need methods to identify suspicious transaction. In recent years, data mining techniques have been able to successfully prevent money laundering and detect credit card fraud. In this study we used k-nearest neighbor technique with association rules to improve accuracy of algorithms for detecting outliers in transactions which is used in credit card in electronic banking system. Finally, the results of proposed method in terms of accuracy and speed have been compared and evaluated with other methods.
Keywords : fraud detection electronic banking knn algorithm association rules