Predicting Student Mental Health with a Data-Driven Approach to Early Intervention and Artificial Intelligence

Abstract

The mental health of students has become a fast-developing area of research, given its impact on academic performance, interpersonal relationships, and overall health. This study utilizes machine learning techniques, specifically Random Forest for classification and K-Means clustering with Principal Component Analysis (PCA), to analyze key factors influencing student mental health, including self-esteem, sleep quality, study load, social support and anxiety levels. A mental health analyzer was developed to sort and analyze student data, identifying distinct groups, including those with high stress and severe anxiety and depression due to academic pressure, those with moderate stress but with some coping capacity, those with stable mental health and minor issues, and those with high well-being, good academic performance, and good social support. The findings emphasize the importance of early intervention, personalized support strategies, and mental health support in educational settings. By integrating machine learning for mental health assessment, this research yields valuable insights to educators and policymakers in designing evidence-based, individualized interventions to improve student well-being and academic performance.