Komparasi Algoritma Klasifikasi Menentukan Kelulusan Mata Kuliah Pada Universitas

Abstract

Abstract. Graduation of the course is a reflection of the seriousness of individual students every semester of the lecture. Students who attend all the learning process in the classroom or face-to-face in each course will have more chance to graduate than the students who are rarely present, some lecturer's evaluation of the graduation of the course are attendance, task, UTS and UAS. From the above problem the writer tries to get the appropriate algorithm to determine the graduation of each student's course, this research will use the data mining classification technique with 3 methods of classification algorithm such as Algorithm C4.5 (decision tree), Naïve Bayes, and Random Forest with label result pass or not pass. Based on the results of tests conducted using the same dataset on the 3 algorithms through comparison to the value of AUC and Confusion Matrix, obtained the value of Area Under Curve (AUC) of 1,000 from Naïve Bayes model, while the largest Accuracy or Confusion Matrix value is in C4 algorithm .5 (decision tree) with a value of 97.78%. Thus, the proper algorithm for this case is the C4.5 Algorithm (decisiontree).

Keywords: Course graduation, data mining classification, Algorithm C4.5 (decision tree), Naïve Bayes, and Random Forest.