OKECHI, AUGUSTINE KENECHUKWU (2026) DEVELOPMENT OF A MACHINE LEARNING-BASED DEPRESSION RISK PREDICTION SYSTEM FOR UNIVERSITY STUDENTS. Other thesis, Godfrey Okoye University, Enugu Nigeria.
![[img]](https://eprints.gouni.edu.ng/style/images/fileicons/text.png) |
Text
Okechi_Augustine_final_report-2.PDF Download (1MB) |
Abstract
University students' depression is a big public health issue, with studies showing high rates that hurt grades, social life, and well-being. Traditional screening methods are usually reactive, costly, and lack access for many students. So, there's a real need for tech-based, early detection tools that can reach more students easily. This project is about an online system that predicts depression risk in university students. It uses the Depression Anxiety Stress Scale (DASS-21) along with machine learning. The system gathers data through twenty-one questions split between depression, anxiety, and stress, and includes five extra lifestyle factors—sleep, exercise, social life, appetite, and academic pressure. The project uses a Random Forest classifier on data from 1,120 students. This method reached 89.3% accuracy and an AUC-ROC of 0.93. That's better than other models like logistic regression, SVM, and K-Nearest Neighbors. The system gives a risk classification of Low, Moderate, or High, along with a confidence score and personalized recommendations. It's built with React.js for the frontend, which offers a responsive multi-step assessment interface. For the backend, they used Python and Flask to create a RESTful API for real-time results. This tool is meant for preliminary screenings only—to help with early referrals and support in university health services—but isn't meant to take the place of a full clinical assessment Keywords: Depression prediction, DASS-21, machine learning, Random Forest, student mental health, Flask, React.js, web-based system.
| Item Type: | Thesis (Other) |
|---|---|
| Subjects: | Q Science > Q Science (General) |
| Divisions: | Faculty of Computing And Information Technology (FACIT) |
| Depositing User: | Cynthia Ugwuoti |
| Date Deposited: | 24 Jul 2026 14:40 |
| Last Modified: | 24 Jul 2026 14:40 |
| URI: | http://eprints.gouni.edu.ng/id/eprint/6009 |
Actions (login required)
 |
View Item |
