Technological Model Based on Machine Learning to Improve the Enrollment Process in Educational Institutions in Northern Lima

Ñiquen R.; Luna J.; Aguirre J. · Proceedings of the LACCEI international Multi-conference for Engineering, Education and Technology (2026) · DOI: 10.18687/laccei2026.1.1.291

Resumen

The enrollment process in public schools in Northern Lima is predominantly managed through manual forms, leading to inefficiencies, risk of information loss and administrative overload. This research proposes a machine learning-based model to optimize the management of enrollment records through risk-based classification, enabling the early identification of cases that may require additional administrative review. A dataset of 4,500 anonymized enrollment records collected between 2024 and 2026 from a public educational institution was used. Three supervised learning algorithms: Random Forest, Logistic Regression and Decision Tree were implemented and evaluated using a time-based 80/20 train-test split where results show that Random Forest achieved an accuracy of 60.6% and the lowest Brier Score (0.238), while Logistic Regression obtained the highest F1-Score (47.2%) and ROC-AUC (60.4%). Based on these results, Random Forest was selected due to its probability calibration and operational applicability. The proposed model enables the classification of enrollment records into risk levels, supporting administrative prioritization and improving decision-making in enrollment management. The study contributes to the digital transformation of the public education system in Peru, with a high potential for scalability at regional and national levels.

Instituciones: Universidad Peruana de Ciencias Aplicadas