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

Abstract

The objective of this research is to undertake a comparative analysis of the performance of the Decision Tree (C4.5) and Naïve Bayes algorithms in the classification of Bank Indonesia scholarship recipients. The CRISP-DM methodology was employed on a data set comprising 416 records of scholarship recipients for the 2023-2024 academic year. The evaluation of the model was conducted using 10-fold cross-validation and the metrics of accuracy, precision, recall, and F-measure. The results demonstrated that the Decision Tree (C4.5) algorithm exhibited superior performance, with an accuracy of 82.70%, precision of 98%, recall of 84.07%, and F-measure of 90.5%, in comparison to the Naïve Bayes algorithm, which demonstrated an accuracy of 82.21%, precision of 97.43%, recall of 83.99%, and F-measure of 90.2%. Although the Decision Tree (C4.5) method requires a slightly longer analysis time, this difference is not significant.