Identifying Credit Card Fraudulent Transactions through Machine Learning Analysis

In the last few years, there has been a sharp rise in credit card fraud transactions. For financial institutions, credit card theft is a severe problem, and reliably detecting fraud can be challenging. An annual study conducted in 2021 found that over 50% of Americans had experienced credit and debit card fraud, and over 13% of those who use these cards often also do so. This indicates that 127 million Americans have experienced credit card theft at some point in the past. Using the traditional way, it is very hard and time-consuming to find this form of fraud in a large database. An effective solution to handle this kind of issue is to create a fraud detection system that is fully automated to identify and categorize such cases using technologies like artificial intelligence (AI) and machine learning. SVM, Random Forest, KNN, XGBoost, and Naïve Bayes are the six supervised machine learning algorithms provided by this study. By using these techniques, a classification model that reliably identifies fraudulent transactions is created. It was discovered that Support Vector Machine was the most accurate model for differentiating between authentic and fraudulent transactions after each of these algorithms was examined.