Rescuing Moroccan Students Before It Is Too Late: Drop Out Early Warning System (Dews)

As a nation navigating its course through education reform, Morocco faces a critical challenge with its student dropout rates. While the primary school dropout rate lingers at 3.6%, there is a significant surge in middle-school, escalating to 14.3%, and a striking 10.4% in high school (as of 2019). Precise identification of students vulnerable to academic discontinuation offers an opportunity for proactive remedial intervention, enabling schools to orchestrate timely preventive measures. This study ventures into the realm of data mining, employing its techniques to predict academic dropouts. It harnesses an expansive array of academic, demographic, and socio-economic student data to equip the predictive model. The objectives of this research are twofold: (1) identifying the pivotal data features that effectively encapsulate the risk factors leading to school dropout and (2) deploying machine learning algorithms to forecast student dropout, ultimately assisting in the early detection of students who require supportive

Intervention. Our comparison of various machine learning methodologies provides an insight into their respective efficacies in identifying students at risk. Encouragingly, our findings reveal an ability to pinpoint 84% of potential dropouts by merely scrutinizing 19% of the dataset. Conclusively, we discern that unauthorized absences, Grade Point Average (GPA), and class rank serve as crucial indicators in predicting school dropout. This research illuminates a potential pathway for the implementation of predictive data science in the education sector, potentially reducing dropout rates and fostering academic success in Morocco.