<![CDATA[Scholarship selection is a critical process in higher education that requires objective, fair, and efficient evaluation of applicants based on academic and socio-economic criteria. However, manual assessment methods are often vulnerable to bias, inconsistency, and administrative inefficiencies, which may affect the transparency and quality of decision-making. This study compares the performance of three supervised machine learning algorithms—C4.5 Decision Tree, Naive Bayes, and K-Nearest Neighbor (KNN)—for scholarship recipient classification. The dataset consisted of 1,500 student records obtained from the KelasAI repository and included ten predictor attributes, namely Grade Point Average, Parental Income, Academic Semester, Family Dependents, Organizational Involvement, Academic Achievement, Regional Origin, Scholarship Type, National Examination Score, and Economic Status. The target variable was categorized into Accepted and Rejected classes. Experiments were conducted using RapidMiner Studio with 10-fold stratified cross-validation to ensure reliable model evaluation. The results showed that Naive Bayes achieved the best performance, with 81.6% accuracy, 81.8% precision, and 81.3% recall, outperforming C4.5 and KNN. These findings demonstrate the potential of machine learning to support more transparent and data-driven scholarship selection processes.]]>
Copyrights © 2026
