|
Enhancing Prediction of Student Learning Success in Higher Education using Deep Learning |
|---|---|
| รหัสดีโอไอ | |
| Creator | Chayaporn Kaensar |
| Title | Enhancing Prediction of Student Learning Success in Higher Education using Deep Learning |
| Contributor | Worayoot Wongnin |
| Publisher | Faculty of Science, Ubon Ratchathani University |
| Publication Year | 2567 |
| Journal Title | Journal of Science and Science Education |
| Journal Vol. | 7 |
| Journal No. | 1 |
| Page no. | 21-36 |
| Keyword | Student performance prediction, Deep learning, Web application, Django, Graduation rate |
| URL Website | https://so04.tci-thaijo.org/index.php/JSSE |
| Website title | Journal of Science and Science Education |
| ISSN | ISSN 2697-410X |
| Abstract | Tackling the global challenge of student attrition and underperformance that widely affect educational quality is a huge challenge. This is a very significant issue in Thailand. Thus, the current research analyze data from 51,106 students and 11 faculties of Ubon Ratchathani University in Thailand for the 2011-2021 academic years. This was done to develop enhanced predictions of academic success utilizing two deep learning methods (Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN)), each fine-tuned for maximum accuracy. The predictive models were scored using the Confusion Metrics and AUC-ROC. The experimental results show that, our research stands out for its robust predictive power, achieving accuracies between 81% to 98%. Notably, RNN achieved high accuracy in all faculties. Additionally, leveraging Django, Python, and PostgreSQL, we propose a research methodology and develop a web application that operationalizes our findings to stakeholders for effectiveness. It has the capability of providing a group of users with a practical tool enabling individual students to forecast academic outcomes, teachers to identify at-risk students early, and faculty as well as, university staff to support informed decision-making thereby improving educational strategies and outcomes. This comprehensive methodology illuminates effective predictive models and system tools driving academic achievement in our university and other higher education institutes. |
