Credit card fraud detection system using smote technique and whale optimization algorithm

Abstract

Credit cards are now being widely used all over the world for transactions, irrespective of the geographical boundaries. Therefore the range of theft and fraud transactions have widely increased. In order to detect both the fraud and non-fraud transactions Credit card fraud detection system (CCFD) is used. Credit card fraud detection system is proposed using machine learning techniques. The two main important algorithm techniques used in this system are whale optimization algorithm(WOA) and Smote(synthetic minority oversampling technique). The Smote technique is used to solve Class imbalance problem. The Whale optimization algorithm comprises mainly of three operators which are used to stimulate the search for prey, encircling prey and bubble-net scratch around the behaviour of humpback whales. It is also used to increase the efficiency of the credit card fraud detection system. The Smote technique is used to solve Class imbalance problem. Thus by using the SMOTE technique and Whale optimization algorithm credit card fraud detection system solves the problem of data imbalance, reliability, data optimization, and improvises the convergence speed.