PERBANDINGAN METODE ALGORITMA C4.5 DAN NAÏVE BAYES UNTUK MEMPREDIKSI KELULUSAN MAHASISWA

Abstract

This study was conducted to compare the accuracy of two algorithm methods, namely the C4.5 algorithm and Naïve Bayes algorithm on a number of datasets. The sources of datasets used are student data of the Informatics Engineering Study Program (STMIK Palangkaraya) where each dataset has a different amount of data (instances) and number of attributes. Based on the results of the comparison study of the C4.5 and Naïve Bayes Algorithm for Predicting the On-time Graduation of STMIK Palangkaraya Students, the results obtained from the accuracy of the two algorithms show that the accuracy of the C4.5 Algorithm is 90% better than the Naïve Bayes algorithm, which is only 85%. The C4.5 algorithm also gives values on recall and precision of 92% and 94% better than the Naïve Bayes algorithm with values of 86% and 93%, respectively.