Detection of Credit Card Fraud Using Machine Learning Algorithms: Insights from the ‘Credit Card Fraud Detection’ Dataset

Abstract

This study investigates the application of machine learning algorithms for credit card fraud detection using the ‘Credit Card Fraud Detection’ dataset from Kaggle. As electronic transactions continue to rise, so does the prevalence of fraudulent activities, necessitating robust detection systems. The research explores three key algorithms: Logistic Regression, Random Forest, and Gradient Boosting. The methodology involves data preprocessing, exploratory data analysis, and model implementation using Python. Performance metrics including accuracy, precision, recall, and F1 score are used to evaluate and compare the effectiveness of each algorithm in identifying fraudulent transactions. Results indicate that while all models show promising performance, Gradient Boosting demonstrates superior accuracy, precision, and recall compared to Logistic Regression and Random Forest. The study highlights the advantages of ensemble and boosting techniques in handling imbalanced datasets and capturing complex fraud patterns. The research fulfills its objectives by exploring machine learning applications in fraud detection, implementing and evaluating models on a real-world dataset, comparing algorithm performance, and recommending the most effective approach. Gradient Boosting is identifiedas the most suitable algorithm for credit card fraud detection based on its performance metrics.Recommendations include integrating Gradient Boosting intofraud detection systems, regularly updating models to adapt to evolving fraud strategies, and combining machine learning with domain expertise. The study contributes to the ongoing development of more effective fraud detection mechanisms in the financial sector and provides a foundation for further research in applying advanced machine learning techniques to cybersecurity challenges.

DOI - https://doi.org/10.65525/SVUP.9788199651500.2026.135-158