Student Dropout Prediction under Temporal Drift: An Empirical Evaluation of Random Forest Performance Across Academic Years

• José Thiago Holanda de Alcântara Cabral Instituto Federal da Paraíba (IFPB)

DOI: https://doi.org/10.5753/sbie.2026.27238

Resumo

Student dropout is a critical challenge in the Federal Network of Professional, Scientific, and Technological Education. This study evaluates the temporal robustness of dropout prediction using longitudinal data (2017–2024) from the Nilo Peçanha Platform. Using rolling temporal validation with Random Forest and Logistic Regression, predictive performance and classification threshold stability were evaluated. Results revealed substantial variability, with AUC varying from 0.34 to 0.83 and optimal thresholds ranging from 0.01 to 0.54. The findings are consistent with the possible occurrence of concept drift in educational data. Student dropout prediction systems require continuous monitoring and periodic recalibration to maintain their effectiveness.

Palavras-chave: Student Dropout, Educational Data Mining, Random Forest, Concept Drift, Temporal Validation

Referências

Ahmed, R., Fahad, N., Miah, M. S. U., Hossen, M. J., and Bhattacharjee, K. (2026). Hyoptensemble: Custom-weighted soft voting hyperparameter optimization ensemble model, explainable-ai for predicting mental state among university students. Discover Artificial Intelligence, 6(1):136.

Alturki, S., Hulpuş, I., and Stuckenschmidt, H. (2022). Predicting academic outcomes: A survey from 2007 till 2018. Technology, Knowledge and Learning, 27(2):275–307.

Bifet, A. and Gavaldà, R. (2007). Learning from time-changing data with adaptive windowing. In Proceedings of the Seventh SIAM International Conference on Data Mining (SDM), pages 443–448. SIAM.

Breiman, L. (2001). Random forests. Machine Learning, 45(1):5–32.

Cabral, J. T. H. d. A. (2025). Predictive modeling for student retention: Evaluation of machine learning algorithms with temporal validation. In Anais do XXXVI Simpósio Brasileiro de Informática na Educação (SBIE 2025), pages 99–112, Curitiba, PR, Brasil. Sociedade Brasileira de Computação (SBC).

Gama, J., Žliobaitė, I., Bifet, A., Pechenizkiy, M., and Bouchachia, A. (2014). A survey on concept drift adaptation. ACM Computing Surveys, 46(4).

Kotsiantis, S. B. (2012). Use of machine learning techniques for educational purposes: A decision support system for forecasting students' grades. Artificial Intelligence Review, 37(4):331–344.

Lu, J., Liu, A., Dong, F., Gu, F., Gama, J., and Zhang, G. (2022). Learning under concept drift: A review. IEEE Transactions on Knowledge and Data Engineering, 34(6):2494–2513.

Reina Marín, Y., Quiñones Huatangari, L., Cruz Caro, O., Maicelo Guevara, J. L., Alva Tuesta, J. N., Sánchez Bardales, E., and Chávez Santos, R. (2025). Data mining to identify university student dropout factors. Applied Sciences, 15(22):11911.

Roberts, D. R., Bahn, V., Ciuti, S., Boyce, M. S., Elith, J., Guillera-Arroita, G., Hauenstein, S., Lahoz-Monfort, J. J., Schröder, B., Thuiller, W., Warton, D. I., Wintle, B. A., Hartig, F., and Dormann, C. F. (2017). Cross-validation strategies for data with temporal, spatial, hierarchical, or phylogenetic structure. Ecography, 40(8):913–929.

Romero, C. and Ventura, S. (2010). Educational data mining: A review of the state of the art. IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews, 40(6):601–618.

Romero, C. and Ventura, S. (2013). Data mining in education. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 3(1):12–27.

Vaarma, M. and Li, H. (2024). Predicting student dropouts with machine learning: An empirical study in finnish higher education. Technology in Society, 76:102474.

Widmer, G. and Kubat, M. (1996). Learning in the presence of concept drift and hidden contexts. Machine Learning, 23(1):69–101.