Predicting student mental health through entropy-based features and interpretable cross-attention transformer networks.

Social support, sleep hours and stress level are the secondary contributors and are moderately significant which indicates that they are meaningful but not dominant.

also in 2,178 other papers