Machine learning-based predictive modelling of mental health in Rwandan Youth

Globally, mental disorders are a significant burden, particularly in low- and middle-income countries, with high prevalence in Rwanda, especially among survivors of the 1994 genocide against Tutsi. Machine learning offers promise in predicting mental health outcomes by identifying patterns missed by traditional methods. However, its application in Rwanda remains under-explored. The study aims to apply machine learning techniques to predict mental health and identify its associated risk factors among Rwandan youth. Mental health data from Rwanda Biomedical Center, collected through the recent Rwanda mental health cross-sectional study and with youth sample of 5221 was used. We used four machine learning models namely logistic regression, Support Vector Machine, Random Forest and Gradient boosting to predict mental health vulnerability among youth. The research findings indicate that the random forest model is the most effective with an accuracy of 88.8% in modeling and predicting factors contributing to mental health vulnerability and 75 % in predicting mental disorders comorbidity. Exposure to traumatic events and violence, heavy drinking and a family history of mental health emerged as the most significant risk factors contributing to the development of mental disorders. While trauma experience, violence experience, affiliation to pro-social group and family history of mental disorders are the main comorbidity drivers. These findings indicate that machine learning can provide insightful results in predicting factors associated with mental health and confirm the role of social and biological factors in mental health. Therefore, it is crucial to consider biological and social factors particularly experience of violence and exposure to traumatic events, when developing mental health interventions and policies in Rwanda. Potential initiatives should prioritize the youth who experience social hardship to strengthen intervention efforts.