Improving Classification Performance in Credit Card Fraud Detection by Using New Data Augmentation

In many industrialized and developing nations, credit cards are one of the most widely used methods payment for online transactions. Credit card invention has streamlined, facilitated, enhanced internet It has, however, also given criminals more opportunities to commit fraud, which raised rate fraud. fraud a concerning global impact; businesses ordinary users have lost millions US dollars as result. Since there is large number transactions, organizations rely heavily on applying machine learning techniques automatically classify or identify fraudulent As performance greatly depends quality training data, imbalance in data not trivial issue. general, only small percentage transactions presented data. This affects classifiers. order deal with rarity occurrences, this paper investigates variety augmentation address imbalanced problem introduces new model, K-CGAN, detection. A main classification then evaluate techniques. These results show that B-SMOTE, SMOTE highest Precision Recall compared other methods. Among those, K-CGAN F1 Score Accuracy.