Towards Transparent Mental Health Insights: An Explainable AI Model for Career-Related Depression and Anxiety Among University Students Using Structured Data

The study presents an Explainable AI (XAI) framework designed to identify early indicators of career-related depression and anxiety among university students, utilizing multimodal data and Federated Learning (FL) for privacy preservation. The model, which incorporates structured behavioral data and facial emotion features through an intermediate fusion neural network with attention mechanisms, achieved an F1-score of 89.12%, accuracy of 92.08%, and precision of 91.88% on the Student Mental Health Survey dataset. This research highlights the potential for scalable and interpretable AI systems in mental health pre-diagnosis, offering insights into behavioral markers that can inform student support services.