CREDIT CARD FRAUD DETECTION USING MACHINE LEARNING APPROACH

Abstract

Payment cards offer a simple and convenient method for making purchases. Generally credit card fraud activities can happen in both online and offline. But in today’s world online fraud transaction activities are increasing day by day. It is therefore crucial to implement mechanisms that can detect the credit card fraud. Features of credit card frauds play important role when machine learning is used for credit card fraud detection, and they must be chosen properly.So in order to find the online

fraud transactions various methods have been used in existing system.In this proposed project de- signed a model to detect the fraud activity in credit card transactions. This system can provide most

of the important features required to detect illegal and illicit transactions.As technology changes constantly it is becoming difficult to track the behavior and pattern of criminal transactions.

The algorithms such as : Random Forest and Extreme Gradient Boosting “XGBoost”. This algo- rithms is based unsupervised learning algorithm. After classification of data set a confusion matrix

is obtained. The performance of the algorithm is evaluated based on the confusion matrix.