Increasing online trading activities or e-commerce has become a trend today. As a result the most common crime is credit card fraud or carding. There are approximately 1,000 cases of fraud in one million transactions so that data is collected in the form of datasets of credit card fraud risk. In some cases, minority classes are more important to identify than the majority class as in the case of credit card transactions. In this study to deal with the problem of class imbalances on credit card fraud risk datasets, the proposed resampling method is the Tomek-Link and SMOT data level with the C5.0 classification model. This research was conducted to improve the accuracy of AUC in the C5.0 classification algorithm model. The results showed that the proposed method was able to increase the AUC value of 0.134 compared to without the resampling method.
Copyrights © 2022