Credit Card Fraud Detection Using Machine Learning

Abstract

Credit card fraud remains a significant challenge in the modern financial landscape, causing substantial economic losses to individuals and institutions. This study presents a machine learning-based approach to proactively identify fraudulent credit card transactions. To improve detection accuracy and minimize false positives, multiple supervised learning algorithms including Logistic Regression, Decision Tree, Random Forest, Support Vector Machine (SVM), and K-Nearest Neighbors (KNN) were implemented and evaluated. The study makes use of a publicly accessible credit card transaction dataset, validating its findings with preprocessing methods and performance metrics like accuracy, precision, recall, and F1-score. A comparative analysis highlights the strengths and weaknesses of each model in handling imbalanced and high-dimensional data. The results demonstrate that ensemble methods like Random Forest offer robust performance in identifying anomalous transaction patterns. The proposed system integrates these models into an admin-controlled platform for real-time detection, making it a scalable and effective solution for fraud mitigation in financial systems