A Comparative Analysis of Supervised Classifiers for Detecting Credit Card Frauds
C P Shilpa, A. H. Shanthakumara · 2022 International Conference on Computer Communication and Informatics (ICCCI) · 2022
Nowadays, Credit Card Frauds are one of the major fraudulent activities due to the vital increase of online payments. Credit Card Fraud generally happens when the card is stolen for any of the unauthorized purposes or even when the Credit Card information is used by fraudster for his use. Hence, we need a system which can notify when a transaction happens to be fraud. Credit Card Fraud Detection System detects given a transaction whether it is fraud or not. This paper mainly focuses on how frauds can be analyzed using Machine Learning approaches. The main objective of this paper focused on methods of handling imbalance in dataset of Credit Card transactions. Later, Credit Card transactions dataset is used on XGBoost and Random Forest Classifier. The results of the XGBoost and Random Forest Algorithms are compared based on the accuracy, precision, recall, and F1- score. The Algorithm that has higher F-1 score is considered to be the best algorithm that is used to detect the fraud.