<![CDATA[The increasing number of students at Universitas Jambi each year is not accompanied by a proportional graduation rate, making it necessary to analyze student data to provide strategic solutions. This study aims to apply data mining techniques to student data to assist academic advisors in predicting students' graduation status and providing early warnings to help students complete their studies on time, thereby minimizing graduation delays. The data used in this study consists of alumni records from the Faculty of Science and Technology at Universitas Jambi from 2019 to 2024, which have undergone a data cleaning process. The classification method used is the Decision Tree algorithm with Chi-Square feature selection to enhance model accuracy. The results indicate that applying feature selection to the Decision Tree algorithm improves the accuracy to 80.00%, compared to 78.57% without feature selection. Additionally, the precision increased from 86.82% to 84.41%, recall improved from 86.72% to 92.24%, and F1-score rose from 86.77% to 88.13%. These findings suggest that feature selection significantly contributes to enhancing the classification model’s performance in predicting student graduation at Universitas Jambi, particularly by improving recall, which reflects the model’s ability to more accurately identify students who graduate on time.]]>
Copyrights © 2025
