Depression Prediction System Using Machine Learning for Student Mental Health Analysis

goal of this research work is to build a ML-based Depression Prediction System that examines the mental health of student and helps predict depression risk at an early stage. This system is developed by extracting some vital academic, behavior, psychological, and life-style based features like Academic pressure, Sleep duration, Study satisfaction, Financial stress, Dietary behavior, Suicidal thoughts, and Family history of mental illness to predict depression among students.</b> The various ML algorithms like Logistic Regression, SVM, KNN, Random Forest, Gradient Boosting and Extra Trees Classifier were applied and tested. In the trained models, ensemble learning algorithms produced better prediction accuracy and performance. Data quality of the given features was improved by applying some preprocessing techniques.    The GUI which was built using Gradio has make the system much interactive and convenient. Through this UI a user can load the dataset, train the model, visualize the result and make prediction manually. Our system can identify whether a student have Low, Medium and High depression risk level and gives better visualization to user about student’s mental health. Our study concluded thatML can be an efficient approach to the depression risk detection and analysis for mental health at university level.    Although the prediction is no way substitute the medical and clinical depression diagnosis and treatment, it can act as a support screening tool for prevention and intervention programs. More possible improvements to this system can be deep learning model, deployment on cloud and web, real-time monitoring, mobile application implementation, and expansion using big data.

[[1] World Health Organization, Depression and Mental Health Report, Geneva, Switzerland: WHO, 2023. [2] J. Smith and P. Kumar, “Machine Learning Techniques for Mental Health Prediction,” International Journal of Artificial Intelligence, vol. 12, no. 3, pp. 45–58, 2022. [3] T. Brown, A. Wilson, and R. Lee, “Student Depression Analysis Using Machine Learning Algorithms,” IEEE Access, vol. 10, pp. 11234–11245, 2022. [4] H. Lee and S. Park, “Predictive Modeling of Depression Using Behavioral and Psychological Features,” Journal of Healthcare Informatics Research, vol. 5, no. 2, pp. 89–102, 2021. [5] R. Johnson, “Impact of Sleep and Academic Stress on Student Mental Health,” Journal of Educational Psychology, vol. 18, no. 1, pp. 21–35, 2020. [6] M. Patel and A. Sharma, “Ensemble Machine Learning Methods for Depression Prediction,” in Proc. International Conference on Data Science and Artificial Intelligence, 2023, pp. 155–160. [7] Y. Zhang, “Explainable AI for Mental Health Prediction Systems,” Springer Journal of AI Applications, vol. 8, no. 4, pp. 211–225, 2022. [8] S. Gupta and N. Rao, “Classification Techniques for Student Mental Health Prediction,” International Journal of Computer Applications, vol. 176, no. 12, pp. 10–18, 2021. [9] K. Anderson and L. Martin, “Data Preprocessing Methods in Healthcare Machine Learning Systems,” Journal of Data Science and Analytics, vol. 14, no. 2, pp. 95–108, 2020. [10] P. Verma and S. Karthik, “Depression Risk Detection Using Ensemble Learning Approaches,” International Journal of Advanced Computer Science and Applications, vol. 13, no. 5, pp. 233–241, 2022.