Description

In this project, I developed a machine learning model to predict student dropouts in Morocco’s middle and high schools. By analyzing data from over 336,000 students in the Fes-Meknes region (2015-2019), I identified key indicators such as unauthorized absences, GPA, and class rank that significantly predict dropout risk. The model I built correctly identified 84% of potential dropouts, helping educators intervene early. This work contributes to reducing dropout rates and promoting academic success through data-driven decision-making in the Moroccan education system.

Poster presentation

Poster presentation video

Scientific paper

Link to Stanford article

Github Repo