Building a Data-Oriented Model for Credit Card Fraud Detection Using Synthetic Data

• Alexandre C. B. dos Santos Cefet/RJ
• Roger de S. Passos Cefet/RJ
• Luis Domingues T. J. Tarrataca Cefet/RJ
• Douglas de O. Cardoso University of Porto
• Diego B. Haddad Cefet/RJ
• Felipe da R. Henriques Cefet/RJ

DOI: https://doi.org/10.5753/sbseg.2024.241488

Abstract

Credit card transaction fraud is a global challenge, resulting in significant financial losses. This work proposes a synthetic data simulator for transactions to replicate the dynamics of real-world data. These data were used to create models based on classification algorithms and anomaly detection, capable of identifying fraudulent transactions. Challenges such as sequential modeling, context change, delayed feedback, and data peculiarities were addressed. The Random Forest algorithm stood out, detecting 76.7% of frauds with a precision of 96.4%.

References

(2022). Mais de 50% das tentativas de fraude são no segmento de bancos e cartões, aponta serasa experian. Disponível em: [link]. Acesso em: 01/06/2024.

(2022). Nilson report. Disponível em: [link]. Acesso em: 01/06/2024.

(2023). Mais de 140 mil cartões foram roubados no brasil e vendidos na ’dark web’ em 2023, diz pesquisa. Disponível em: [link]. Acesso em: 01/06/2024.

(2023). Report on card fraud in 2020 and 2021. Disponível em: [link]. Acesso em: 01/06/2024.

Cerqueira, V., Torgo, L., and Mozetič, I. (2020). Evaluating time series forecasting models: An empirical study on performance estimation methods. Machine Learning, 109:1997–2028.

Davis, J. and Goadrich, M. (2006). The relationship between precision-recall and ROC curves. In Cohen, W. W. and Moore, A. W., editors, Machine Learning, Proceedings of the Twenty-Third International Conference (ICML 2006), Pittsburgh, Pennsylvania, USA, June 25-29, 2006, volume 148 of ACM International Conference Proceeding Series, pages 233–240. ACM.

Gama, J., Zliobaite, I., Bifet, A., Pechenizkiy, M., and Bouchachia, A. (2014). A survey on concept drift adaptation. ACM Comput. Surv., 46(4):44:1–44:37.

Le Borgne, Y.-A., Siblini, W., Lebichot, B., and Bontempi, G. (2022). Reproducible Machine Learning for Credit Card Fraud Detection - Practical Handbook. Université Libre de Bruxelles.

Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., and Duchesnay, E. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12:2825–2830.

Wong, N., Ray, P., Stephens, G., and Lewis, L. (2012). Artificial immune systems for the detection of credit card fraud: an architecture, prototype and preliminary results. Inf. Syst. J., 22(1):53–76.