USING MACHINE LEARNING PRACTICES TO DETECT FRAUD AND RISK MANAGEMENT

This paper presents a machine learning (ML) approach to detect credit card fraud and manage related risks in transactions. Traditional fraud detection systems struggle to adapt to evolving fraudulent patterns, leading to financial losses. ML techniques offer improved accuracy and efficiency in fraud detection. The proposed framework includes data preprocessing, feature engineering, model training, and evaluation. Key features such as transaction amount and location are extracted for analysis. ML models, including neural networks, decision trees, and logistic regression, are trained and evaluated using precision, recall, F1 score, and AUC-ROC. Cross-validation is employed for hyperparameter tuning, and a hybrid approach integrates supervised learning with anomaly detection to reduce false positives. Experiments use synthetic data to ensure privacy. This study contributes to safer and more transparent credit card transactions through advanced ML practices, enhancing fraud detection accuracy.