Predictability: Advancing Mental Health Modeling for Students

This study explores the development of a machine learning-based approach to proactively manage student mental health in higher education. Through these advanced analytical techniques, key mental health determinants were identified, integrating demographic, academic, socioeconomic, and behavioral factors. A Random Forest model, optimised via hyperparameter tuning, demonstrated superior predictive performance with an accuracy of 91% effectively classifying at-risk students. The determinants translate these findings into actionable insights, offering educators a comprehensive tool for timely interventions. By bridging technology and mental health management, this research underscores the transformative potential of predictive frameworks in fostering inclusive, supportive academic environments. This work advances the discourse on student mental health, emphasising data-driven strategies for early detection and intervention to improve educational outcomes.