AAI_2025_Capstone_Chronicles_Combined

MENTAL HEALTH RISK DETECTION USING ML​

​

​

​

​

19

difficulty, family history, and social withdrawal, many of which showed strong statistical associations with the target variable. Even after removing dominant clustering features, the model sustained strong performance, confirming that meaningful and generalizable patterns existed in the remaining data. This evidence validates our hypothesis and demonstrates the predictive value of the selected survey features.

Figure 5 Mental Health Risk Model: Infrastructure and Workflow

Note. An end-to-end pipeline illustrating user input, data preprocessing, model evaluation, and final interpretation by clinicians.​

219