In this paper we examine the development and implementation of a credit card fraud detection system using machine learning and big data technologies on the Google Cloud® platform. With global financial losses from credit card fraud exceeding £28.06 billion in 2020, this paper addresses the need for advanced solutions to enhance fraud detection. Informed by the need to combat sophisticated fraud, the paper explores advanced techniques to improve detection rates. A literature review reveals limitations in traditional fraud detection approaches, which struggle to adapt to changing fraud patterns. While machine learning algorithms show promise, there is a gap in integrating these algorithms with big data technologies for robust frameworks. This paper combines advanced machine learning techniques to develop an effective fraud detection system. By investigating various algorithms and their integration with big data platforms, the study aims to create a scalable solution for detecting fraudulent transactions. The methodology involves using Google Cloud® Platform, Google Drive, and Google Colab to implement and evaluate eight models, including logistic regression, random forest, XGBoost, LightGBM, SVM, feedforward neural network, CNN, and decision tree. Each algorithm's performance is evaluated through metrics like accuracy, precision, recall, F1-score, and ROC-AUC. Findings reveal that ensemble methods and CNNs are top performers, showcasing high accuracy and balanced precision and recall. This approach offers financial institutions a scalable and adaptive solution, reducing financial losses and enhancing transaction security. Future work includes continuous model monitoring, real-time processing, feature engineering, and incorporating external data sources to enrich the dataset.

Authors: T.K. Olaniyi, S. Ihsan

Published in: International Conference for Internet Technology and Secured Transactions (ICITST-2024)

• Date of Conference: 4-6 November 2024
• DOI: 10.20533/ICITST.2024.0028
• ISBN: 978-1-913572-76-1
• Conference Location: St Anne’s College, Oxford University, UK