Prediction Model for Retention in the First Year of Higher Education: the Case of Faculty of Arts and Letters

A major challenge for higher education institutions in Burkina Faso is student retention, particularly in the first year. The rise of predictive analytics and big data tools presents an opportunity to improve the guidance of new high school graduates through the Campusfaso platform. In this paper, we propose a tool for predicting student retention in the first year of Letters programs in higher education in Burkina Faso using boosting algorithms. Deploying a machine learning model will help high school graduates choose the most suitable higher education programs. This study was conducted on a sample of 18,812 student records from six universities in Burkina Faso in the Faculty of arts and letters. The Data analysis shows that retention is higher in the first year at University4, University5, and University6. Boosting algorithms were used for implementation and model training. The results demonstrate the suitability of the model, with an accuracy rate of 72%. Hyperparameter tuning was performed using GridSearchCV, and model performance was evaluated using metrics such as accuracy, precision, recall, F1–score, and ROC curves.