Prediction of IPVA Default in the State of Ceará Using Machine Learning

• Maria Inês V. Silva Federal University of Ceará (UFC) / Secretariat of Finance of the State of Ceará (SEFAZ-CE) http://orcid.org/0000-0002-0745-5054
• Francisco Victor da S. Pinheiro Federal University of Ceará (UFC) https://orcid.org/0000-0002-5693-0499
• César Lincoln C. Mattos Federal University of Ceará (UFC) https://orcid.org/0000-0002-2404-3625
• José Maria da S. Monteiro Filho Federal University of Ceará (UFC) https://orcid.org/0000-0002-5583-6070
• Rossana M. C. Andrade Federal University of Ceará (UFC) https://orcid.org/0000-0002-0186-2994

DOI: https://doi.org/10.5753/sbbd_estendido.2024.243911

Abstract

For Brazilian states and municipalities, the Motor Vehicle Ownership Tax (IPVA) is the second most important source of revenue. Compliance with the collection of IPVA depends on different factors, such as the country’s economy, the market value and residence of the vehicles, among other factors. Predicting whether taxpayers will be compliant or not in relation to the payment of IPVA can provide subsidies that help governments to develop public policies, planning tax actions and directing campaigns to encourage timely tax payment. In this work, we conducted a series of experiments aiming to build an efficient solution to the problem of classifying taxpayers in terms of their compliance with the IPVA payment. Real data referring to the IPVA of the State of Ceará was used from 2019 to 2023. In total, four classification algorithms were explored to classify taxpayers into two groups: compliant and non-compliant. The best results achieved an F1 score of 0.86 proving the viability of the proposed solution.

Keywords: Prediction, Default, IPVA

References

Akinrinola, O., Addy, W. A., Ajayi-Nifise, A. O., Odeyemi, O., and Falaiye, T. (2024). Application of machine learning in tax prediction: A review with practical approaches. Global Journal of Engineering and Technology Advances, page 102–117.

Amarasinghe, K., Rodolfa, K. T., Lamba, H., and Ghani, R. (2023). Explainable machine learning for public policy: Use cases, gaps, and research directions. Data 38; Policy, 5:e5.

Brasil (1988). Constituição da República Federativa do Brasil. Acesso em: 25 de maio de 2024.

Faceli, K. (2011). Inteligência artificial: uma abordagem de aprendizado de máquina. Grupo Gen - LTC.

Governo do Estado do Ceará (2023). Sefaz divulga tabela do IPVA 2024, que apresenta redução média de 4,59Acesso em: 28 de maio de 2024.

Hastie, T., Tibshirani, R., Friedman, J. H., and Friedman, J. H. (2009). The elements of statistical learning: data mining, inference, and prediction, volume 2. Springer.

Lima, M. S. M. and Delen, D. (2020). Predicting and explaining corruption across countries: A machine learning approach. Government Information Quarterly, 37(1):101407.

Silva, D., Carvalho, S. T., and Silva, N. (2022). Comparative analysis of classification algorithms applied to circular trading prediction scenarios. In Kő, A., Francesconi, E., Kotsis, G., Tjoa, A. M., and Khalil, I., editors, Electronic Government and the Information Systems Perspective, pages 95–109, Cham. Springer International Publishing.