Prediksi Kelulusan Mahasiswa Dengan Metode Naive Bayes dan Artificial Neural Network : Studi Kasus Fakultas Teknik UNIS Tangerang

Abstract

The faculty of engineering has 4(four) studies programs namely, informatics engineering, civil engineering, industrial engineering, chemical engineering. The number of lecturers and students owned by the Faculty of Engineering based on PDDikti Year1 2019/2020 reporting data is 41 permanent lecturers and 750 students. The problems faced by the Faculty of Engineering UNIS Tangerang include the low percentage of students who graduate on time compared to students who graduate not on time. In the 2015/2016 graduation year, only 30% of students passed on time, the rest did not graduate on time. This study aims to assist the Faculty of Engineering in predicting student graduation, so that it can be anticipated earlier. This research uses the attributes of total credits, 1st semester IP, 2nd semester IP, 3rd semester IP, 4th semester IP. The methods used in this research are Naïve Bayes and Artificial Neural Network. The data used in this study used 330 records of students who graduated in 2012-2016. The results of the accuracy obtained after testing with the system using 20% data testing obtained an accuracy of 63.63%, 71.05% precision, 67.5% recall, and 62.6% AUC.