AI-Based Application to Predict Student Dropout in the National Learning Service SENA

Abstract

We created a web application for the National Learning Service (NLS) in Colombia to identify students who are at risk of dropping out. This application helps the Student Welfare Department offer support to those who need it. We collected data on socio-economic variables that contributed to academic dropout and used it to develop the software. We explored different artificial intelligence techniques in the Machine Learning area and selected the best algorithm to integrate into the software. The resulting software has improved processes at NLS and can benefit other higher education institutions.