ML-BASED FRAUD DETECTION FOR SECURING CREDIT CARD

Abstract

Credit cards are used a lot for online payments because they are fast and easy to use. But as more people use credit cards, fraud has also increased. This fraud causes big money losses for both users and banks. In this study, the main goal is to detect credit card fraud using machine learning methods. We used public data for this study, which had challenges like very few fraud cases, changing fraud patterns, and many false alerts.Many earlier studies have used machine learning methods like Decision Trees, Random Forest,LSTM, CNN and stacking classifier, these methods give high accuracy. So, we tested machine learning methods to improve the results., we used machine learning models, which gave better results than before. Then, we applied five models which are machine learning algorithms.

We tried different numbers of layers, training rounds (epochs), and updated models to get the best performance. Our final model gave very high results.