Plataforma preditiva de combate à evasão escolar para o Ensino Fundamental: construção do modelo preditivo
Resumo
A evasão escolar no Ensino Fundamental costuma ser antecedida por desengajamento, infrequência e baixo desempenho nem sempre tratados precocemente. Este artigo apresenta o modelo preditivo de uma plataforma de combate à evasão, baseada em IA e mineração de dados, para identificação precoce de risco. A solução combina pré-processamento, aprendizado de máquina, classificação de risco e dashboards, seguindo o CRISP-DM. Foram comparados cinco modelos em um conjunto de 15.193 registros e 73 colunas (84,36% permanência, 15,64% evasão). O XGBoost com SMOTE apresentou o melhor equilíbrio: acurácia de 0,912, F1-score de 0,718, precisão de 0,720 e recall de 0,716. Os resultados indicam viabilidade técnica, com validação ainda pendente.
Palavras-chave:
Evasão escolar, Mineração de dados educacionais, Aprendizado de máquina, Sistemas de alerta precoce, Ensino Fundamental
Referências
Baker, R. S. J. d. e Yacef, K. (2009). The state of educational data mining in 2009: A review and future visions. Journal of Educational Data Mining, 1(1):3–17.
Barker, C. G. e Siddiqui, N. (2026). Attendance trajectories as early predictors of school dropout: Evidence from a national cohort in chile. International Journal of Educational Development.
Bowers, A. J., Sprott, R., e Taff, S. A. (2013). Do we know who will drop out? a review of the predictors of dropping out of high school: Precision, sensitivity, and specificity. The High School Journal, 96(2):77–100.
Buckingham Shum, S. e Ferguson, R. (2012). Social learning analytics. Journal of Educational Technology & Society, 15(3):3–26.
Chawla, N. V., Bowyer, K. W., Hall, L. O., e Kegelmeyer, W. P. (2002). SMOTE: Synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, 16:321–357.
Chen, T. e Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 785–794, San Francisco, CA. ACM.
El Jihaoui, M., Abra, O. E. K., e Mansouri, K. (2025). Predicting and explaining middle-school dropout risk on imbalanced data. E3S Web of Conferences, 680:00073.
Gallardo, C. D. C., Chulluncuy, L. V. T., e Quintana, C. A. A. (2025). Web application based on machine learning for the prediction and prevention of low academic performance in primary school students in the district of matara, cajamarca. In Proceedings of the IEEE International Conference on Big Data.
Han, Y. et al. (2025). Distinguishing characteristics of out-of-school adolescents in south korea: A machine learning approach. International Journal of Social Welfare. Publicado online em outubro de 2024.
Hidalgo-Hidalgo, K. J., Huaricancha-Estrella, R., Abarca-Torres, A. R., Huaman-Alegria, M., Espino-Oncebay, P., e Rodriguez-Baca, L. S. (2025). Predicting school dropout rates in primary education using artificial intelligence techniques. In International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME 2025).
IBGE (2025). Pesquisa nacional por amostra de domicílios contínua: Educação 2024. Technical report, IBGE, Rio de Janeiro. Módulo anual de educação, divulgado em junho de 2025.
INEP (2025). Censo escolar da educação básica 2024: Resumo técnico. Technical report, INEP, Brasília.
Psyridou, M., Prezja, F., Torppa, M., Lerkkanen, M.-K., Poikkeus, A.-M., e Vasalampi, K. (2024). Machine learning predicts upper secondary education dropout as early as the end of primary school. Scientific Reports, 14.
Rezk, S. S. e Selim, K. S. (2024). Ensemble pruning for predicting school dropouts: A machine learning approach. Journal of Computational Social Science, 7(2):1555–1597.
Rodríguez, P., Villanueva, A., Dombrovskaia, L., e Valenzuela, J. P. (2023). A methodology to design, develop, and evaluate machine learning models for predicting dropout in school systems: The case of chile. Education and Information Technologies, 28:10103–10149.
Saito, T. e Rehmsmeier, M. (2015). The precision-recall plot is more informative than the roc plot when evaluating binary classifiers on imbalanced datasets. PLOS ONE, 10(3):e0118432.
Selim, K. S. e Rezk, S. S. (2023). On predicting school dropouts in egypt: A machine learning approach. Education and Information Technologies, 28(7):9235–9266.
Shearer, C. (2000). The CRISP-DM model: The new blueprint for data mining. Journal of Data Warehousing, 5(4):13–22.
Barker, C. G. e Siddiqui, N. (2026). Attendance trajectories as early predictors of school dropout: Evidence from a national cohort in chile. International Journal of Educational Development.
Bowers, A. J., Sprott, R., e Taff, S. A. (2013). Do we know who will drop out? a review of the predictors of dropping out of high school: Precision, sensitivity, and specificity. The High School Journal, 96(2):77–100.
Buckingham Shum, S. e Ferguson, R. (2012). Social learning analytics. Journal of Educational Technology & Society, 15(3):3–26.
Chawla, N. V., Bowyer, K. W., Hall, L. O., e Kegelmeyer, W. P. (2002). SMOTE: Synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, 16:321–357.
Chen, T. e Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 785–794, San Francisco, CA. ACM.
El Jihaoui, M., Abra, O. E. K., e Mansouri, K. (2025). Predicting and explaining middle-school dropout risk on imbalanced data. E3S Web of Conferences, 680:00073.
Gallardo, C. D. C., Chulluncuy, L. V. T., e Quintana, C. A. A. (2025). Web application based on machine learning for the prediction and prevention of low academic performance in primary school students in the district of matara, cajamarca. In Proceedings of the IEEE International Conference on Big Data.
Han, Y. et al. (2025). Distinguishing characteristics of out-of-school adolescents in south korea: A machine learning approach. International Journal of Social Welfare. Publicado online em outubro de 2024.
Hidalgo-Hidalgo, K. J., Huaricancha-Estrella, R., Abarca-Torres, A. R., Huaman-Alegria, M., Espino-Oncebay, P., e Rodriguez-Baca, L. S. (2025). Predicting school dropout rates in primary education using artificial intelligence techniques. In International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME 2025).
IBGE (2025). Pesquisa nacional por amostra de domicílios contínua: Educação 2024. Technical report, IBGE, Rio de Janeiro. Módulo anual de educação, divulgado em junho de 2025.
INEP (2025). Censo escolar da educação básica 2024: Resumo técnico. Technical report, INEP, Brasília.
Psyridou, M., Prezja, F., Torppa, M., Lerkkanen, M.-K., Poikkeus, A.-M., e Vasalampi, K. (2024). Machine learning predicts upper secondary education dropout as early as the end of primary school. Scientific Reports, 14.
Rezk, S. S. e Selim, K. S. (2024). Ensemble pruning for predicting school dropouts: A machine learning approach. Journal of Computational Social Science, 7(2):1555–1597.
Rodríguez, P., Villanueva, A., Dombrovskaia, L., e Valenzuela, J. P. (2023). A methodology to design, develop, and evaluate machine learning models for predicting dropout in school systems: The case of chile. Education and Information Technologies, 28:10103–10149.
Saito, T. e Rehmsmeier, M. (2015). The precision-recall plot is more informative than the roc plot when evaluating binary classifiers on imbalanced datasets. PLOS ONE, 10(3):e0118432.
Selim, K. S. e Rezk, S. S. (2023). On predicting school dropouts in egypt: A machine learning approach. Education and Information Technologies, 28(7):9235–9266.
Shearer, C. (2000). The CRISP-DM model: The new blueprint for data mining. Journal of Data Warehousing, 5(4):13–22.
Publicado
05/10/2026
Como Citar
DE SOUSA, Gabriel Monteiro Cordeiro; MACEDO, Lucas Francisco Alcantara Sales; DA CUNHA, David Carlos Pereira; GOMES, Alex Sandro.
Plataforma preditiva de combate à evasão escolar para o Ensino Fundamental: construção do modelo preditivo. In: SIMPÓSIO BRASILEIRO DE INFORMÁTICA NA EDUCAÇÃO (SBIE), 37. , 2026, Goiânia/GO.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2026
.
p. 2956-2965.
DOI: https://doi.org/10.5753/sbie.2026.28501.
