Volume 15, Issue 6 And S7 (Artificial Intelligence- In Press 2025)                   J Research Health 2025, 15(6 And S7): 0-0 | Back to browse issues page

Ethics code: IR.TUMS.MEDICINE.REC.1402.691

XML Print


Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

Keykha A, Mojtahedzadeh R, Keramatfar A, Taghavi Monfared A, Mohammadi A. Explainable Artificial Intelligence in Evaluating the Impact of Tuition Fees on Students’ Academic Decision-Making: A Case Study. J Research Health 2025; 15 (6)
URL: http://jrh.gmu.ac.ir/article-1-2816-en.html
1- Postdoctoral Researcher at Sharif University of Technology, Sharif Policy Research Institute, Tehran, Iran.
2- Department of E-learning in Medical Education, Center of Excellence for E-learning in Medical Education, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran.
3- NLP Engineer, Stinas company, Tehran, Iran.
4- Department of Educational Administration and Planning, Faculty of Psychology and Education, University of Tehran, Tehran, Iran.
5- Department of E-learning in Medical Education, Center of Excellence for E-learning in Medical Education, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran , aeen_mohammadi@tums.ac.ir
Abstract:   (12 Views)
Background: Tuition fees play a pivotal role in shaping students’ academic decisions, influencing decisions such as dropout, major change, guest enrollment, and institutional transfer. While previous studies have examined general predictors of academic behavior, the specific impact of tuition costs has received limited attention, particularly within the context of health-related higher education in Iran. This study seeks to address this gap by developing predictive models to evaluate how tuition-related financial pressures influence students’ academic decisions in the healthcare education system.
Methods: This study analyzed administrative data from TUMS students (2016–2023) using machine learning models, including logistic regression, ANN, decision trees, and random forests. To address class imbalance, SMOTE and random under-sampling were applied. Models were trained under various hyperparameter configurations. The best-performing model–sampling combinations were selected based on accuracy, precision, recall, and F1-score.
Results: The findings demonstrate that machine learning models can effectively predict key academic decisions such as student dropout and guest enrollment. The selected models achieved high overall accuracy (95–99%) and showed acceptable sensitivity to minority classes, particularly for detecting students of dropout (F1 = 0.37) and guest enrollment (F1 = 0.70). These results highlight the potential of predictive analytics to support early interventions and inform data-driven academic policy planning at the institutional level.
Conclusion: This study introduces a data-driven framework for modeling tuition-related academic decisions within the context of health-related higher education in Iran. By employing machine learning techniques and applying class imbalance correction methods, the research enhances both the precision and depth of analysis. The findings offer practical tools for the early identification of at-risk students and support evidence-based decision-making in healthcare education policy and planning.
Full-Text [DOCX 48 kb]   (4 Downloads)    
Type of Study: Orginal Article | Subject: ● Artificial Intelligence
Received: 2025/06/7 | Accepted: 2025/09/14 | Published: 2025/12/24

Add your comments about this article : Your username or Email:
CAPTCHA

Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

© 2025 CC BY-NC 4.0 | Journal of Research and Health

Designed & Developed by : Yektaweb