PERBANDINGAN ALGORITMA C4.5 DAN NAÏVE BAYES UNTUK KLASIFIKASI PENERIMA BEASISWA BANK INDONESIA SUMATERA SELATAN

This study aims to compare the performance of the Decision Tree (C4.5) and Naïve Bayes algorithms in classifying Bank Indonesia scholarship recipients based on data from the 2023-2024 academic year. The CRISP-DM methodology was applied, with model evaluation conducted using 10-fold cross-validation and metrics such as accuracy, precision, recall, and F-measure. The results indicate that the Decision Tree (C4.5) algorithm outperformed Naïve Bayes, achieving 82.70% accuracy, 98% precision, 84.07% recall, and a 90.5% F-measure. In comparison, Naïve Bayes obtained 82.21% accuracy, 97.43% precision, 83.99% recall, and a 90.2% F-measure. Although the Decision Tree (C4.5) requires slightly longer analysis time, it proved to be more effective for this classification task. This study concludes that Decision Tree (C4.5) is the most suitable method for supporting scholarship selection processes, providing new insights into applying data mining technology to improve selection efficiency and accuracy.