Detecting Credit Card Fraud Using Machine Learning Models
Vangipuram Sravan Kiran, D. Srinivasa Rao, Naini Sudeepthi, Anuhya Srestha Sangars, Vontary Ruthvij Reddy, Neha Yalavarthi · 2024
Many cases of credit card fraud occur often, this resulting in huge losses. A large share of the proliferation of online transactions in recent years has now been contributed to online credit card transactions. Therefore, credit card fraud detection applications are a valuable and needed instrument for banks and other financial institutions. The numerous kinds and categories of substantive forms of fraudulent transactions make them unique in many ways. Several models employing the machine learning algorithm are employed in combating the fraud and a performance evaluation is carried out to establish the best technique. All in all, in this review a detailed guide is provided as how to select the right algorithm. An appropriate performance matrix is utilized to illustrate proposed method. The study selected the Credit Card Fraud Detection dataset. In performing feature selection, the dataset is split into two datasets, training data and test data and Alert the User when the model predicts the transaction as fraud. Policies implemented in the experiment that adopted algorithms include - Naive Bayes, K Nearest Neighbors, Random Forest, Logistic Regression, SVM and Decision trees. Among all machine learning models random forest outperforms all with the detection accuracy of 94.98%