Predicción de la matrícula en Educación Básica Especial en el Perú utilizando el algoritmo Prophet: Predicting enrollment in Special Basic Education in Peru using the Prophet algorithm

Predicción de la matrícula en Educación Básica Especial en el Perú utilizando el algoritmo Prophet

Abstract

The present study aimed to predict national enrollment in Special Basic Education (SBE) in Peru for the year 2025, based on the analysis of historical records from the 2019–2024 period. SBE, as a modality aimed at students with Special Educational Needs (SEN), represents a fundamental pillar in building an inclusive and equitable educational system. For this study, an official database from the Peruvian Ministry of Education (ESCALE) was used, which includes a total of 23,905 records with variables such as year, department, educational level, type of management, and number of enrolled students (TALUMNOS). After filtering and cleaning the data, the Prophet algorithm, specialized in time series modeling, was applied to estimate enrollment by level (Preschool, Primary, and Total Basic) for the year 2025. The results show a stabilization of enrollment around 28,371 students, very similar to the total recorded in 2024, suggesting a possible maturity in the SBE coverage system. The model’s accuracy was validated using statistical metrics, obtaining a coefficient of determination (R²) higher than 0.85 at all levels, and an R² of 0.9999 for the national total. Additionally, the MAE, RMSE, and MAPE values reflect excellent fit and low deviation. This study demonstrates the potential of machine learning models in inclusive educational planning and offers a useful tool for the formulation of public policies aimed at students with special educational needs (SEN).