Credit Card Fraud Detection Using Machine Learning Models: A Performance Comparison
Şahide Şeker, Tuğba Dayıoğlu · 2025
Credit card fraud poses a significant threat to the financial sector. This study aims to compare the performance of machine learning models in detecting credit card fraud. The imbalanced dataset containing credit card transactions was balanced using the Synthetic Minority Over-sampling Technique (SMOTE), and Logistic Regression, Random Forest, XGBoost, and LightGBM models were applied. The models' performances were evaluated using metrics such as accuracy, precision, recall, F1-score, and ROC-AUC. The results indicate that Random Forest and XGBoost exhibit the highest success rates. This study provides important insights into determining the most effective machine learning models for detecting credit card fraud.