Penerapan Algoritma K-Means Clustering pada Sistem Prediksi Kelulusan Tepat Waktu

Abstract

Currently, Dikti sets a good on-time graduation percentage for study program accreditation at around 50% for each class. Therefore, the Head of Study Program requires information on the number of students predicted to graduate on time in the eighth semester later, so that policies can be taken as early as possible if the number is not as expected. The method used in this study is the K-Means Clustering algorithm, where this algorithm will divide student data into two groups (clusters), namely the number of students predicted to graduate on time and those who do not graduate on time. The data set used is student academic data from semester one to semester six with five criteria, namely GPA up to semester six, number of credits graduated up to semester six, number of semesters taken up to semester six, number of leaves up to semester six and type of school origin. The results of this study indicate that the number of students predicted to graduate on time is around 92.84% (311 students) based on the sixth semester student data set totaling 335 students with six iterations on the K-Means Clustering algorithm