Machine learning-based analysis and prediction of factors influencing mental health among children and adolescents in Jiangsu Province

Comprehensive Summary

This study uses machine learning to look at the mental health status of children in the Jiangsu Province. A cross-sectional survey was for study which gathered data from 141,725 students. 57 socio-demographic and behavioral factors were used to measure depression, anxiety, and stress scores. 5 machine learning models (Decision Tree, Naïve Bayes, Radom Forest, K-Nearest Neighbors, and XGBoost). In the results, females displayed higher rates among all conditions. Urban areas showed more mental health disparities compared to suburban regions. It was also found that mental health deteriorated as education level increased (depression jumped from 9.2% in primary school to 21.2% in senior high). From the machine learning models, XGBoost achieved the best predictive performance, having achieved AUC scores of 0.799 for depression, 0.770 for anxiety, and 0.762 for stress. The top risk factors were found to be bullying duration, age, and drinking history. In the discussion, the authors talked about the need for programs to help reduce the risk factors associated with poor mental health.

Outcomes and Implications

The information from this study could be used to develop more sophisticated and personalized interventions by identifying mental health risk factors among a pediatric population. Based on the risk factors found, it emphasizes the need for more anti-bullying programs and age-tailored counseling for adolescents. By using machine learning, students can be identified much earlier and receive treatment quicker.