Machine Learning Approaches to Credit Card Fraud Detection

Abstract: Credit card fraud continues to be a major problem in today's digital world, and it needs to be detected and prevented. With the rise of online shopping, the risk of fraud goes up, so we need to find better ways to stop it. Machine learning algorithms are a promising way to spot fraudulent transactions in large amounts of credit card data. But there are challenges to algorithm accuracy, such as not having enough data, having too few examples of fraud, and fraudsters using new tricks all the time. In this study we have looked how well three common machine learning algorithms—Random Forest, Decision Tree, and Logistic Regression—work at spotting credit card fraud using real-world data from transactions. To balance the uneven distribution of fraudulent and genuine transactions, we use SMOTE. We compare algorithms using precision and recall, which measure their ability to identify fraud accurately. Machine learning models analyse historical transactions, recognizing patterns that distinguish normal and fraudulent behaviour. This approach improves fraud detection for credit card transactions, ensuring financial security in online commerce