|
Analysis on Factors Affecting Student Termination Using Data Mining Techniques A Case of Undergraduate Students, Faculty of Business Administration, Rajamangala University of Technology Thanyaburi |
|---|---|
| รหัสดีโอไอ | |
| Creator | Chalee Jittreephong |
| Title | Analysis on Factors Affecting Student Termination Using Data Mining Techniques A Case of Undergraduate Students, Faculty of Business Administration, Rajamangala University of Technology Thanyaburi |
| Contributor | Salitta Saributr |
| Publisher | Faculty of Management Science Nakhon Pathom Rajabhat University. |
| Publication Year | 2568 |
| Journal Title | Journal of Management Science Nakhon Pathom Rajabhat University |
| Journal Vol. | 12 |
| Journal No. | 1 |
| Page no. | 148-167 |
| Keyword | Student Attrition, Data Mining, Early Warning System, Academic Performance |
| URL Website | https://so03.tci-thaijo.org/index.php/JMSNPRU/issue/view/18024 |
| Website title | https://so03.tci-thaijo.org/index.php/JMSNPRU/index |
| ISSN | 2392-5817 |
| Abstract | Student attrition in higher education is an important issue that affects the efficiency of university management and the quality of graduates. This research aimed to analyze the factors that influence student attrition among undergraduate students at the Faculty of Business Administration, Rajamangala University of Technology Thanyaburi, and to develop a risk prediction model using data mining techniques. The study used secondary data from 5,034 students who enrolled between the academic years 2019 and 2022. The analysis followed the Knowledge Discovery in Databases (KDD) process, which included data cleaning, data transformation, and the FP-Growth algorithm to find association rules. Classification models were created using Decision Tree, Na?ve Bayes, and k-Nearest Neighbors (KNN), and their performance was evaluated using Accuracy, Precision, Recall, F1-Score, and AUC. The results showed that the Na?ve Bayes model achieved the highest AUC value at 1.000, showing a strong ability to identify students at risk. The Decision Tree model gave the highest accuracy at 97.45% and had the advantage of being easy to interpret and apply. The important factors related to student attrition were GPA lower than 2.00, low grades in core courses, repeated course enrollment, and poor academic performance in the first year. These findings can be applied to improve early warning systems and proactive student advising effectively. |
