November 28, 2025ElectronicsOpen Access

Advancing Mental Health Detection Through Transfer Learning and Feature Fusion: Mitigating Data Imbalance in Large Transformer Models

View Full Paper

Ask AI

Bookmark

Authors

Discussion

An approach integrating transfer learning improves accuracy in detecting anxiety, depression, and anorexia, addressing class imbalance.

Model accuracy improved significantly by integrating gradient boosting with transfer learning techniques.
Performance metrics for anxiety, depression, and anorexia were enhanced, including notable increases in accuracy and recall rates.
Analysis included confusion matrices to assess model performance on complex emotional data effectively.
Findings suggest innovative solutions for class imbalance may aid future mental health detection frameworks.

Yang et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark