Student performance and health management technology based on GPA model and psychological data mining

In the information age, managing student performance and mental health is critical for educational development.Addressing current analytical limitations, this study proposes an intelligent management framework grounded in data analytics and knowledge integration.The system utilises a four-layer architecture integrating distributed databases and deep learning.Specifically, performance prediction uses a GPA-based model, while an improved two-stream encoder network (TSEN) enables mental health monitoring.Experimental results demonstrate the performance model achieves an accuracy of 0.947, a 0.949 F1 score, and an area under the curve of 0.963 for top predictions.For mental health analysis, using eight influencing factors yields 87.5% sensitivity and 90.1% recognition accuracy, with a Matthews correlation coefficient of 0.962 over a 12-week sequence.These results confirm that the approach effectively integrates academic and psychological data, significantly enhancing analytical accuracy and providing robust decision support for student development and educational management.