Penerapan Data Mining untuk Memprediksi Mahasiswa Drop Out Menggunakan Support Vector Machine

Abstract

Students are one important asset in a university, especially in private universities. The trends of acquired new students for private universities in Indonesia has declined and made new-enrollment process more challenging than ever before. The trend of a late forfeit of registration makes it difficult for one university to predict the number of their new intake students in every new academic year. This research attempts to predict whether one prospective student will likely to continue their study or not, using a data mining method called ID3 algorithm. Set of rules will be created as a basis to be compared with new data which at the end will result in a prediction of whether a student will continue their study(registration process) or not. The output of this DSS System will be implemented in the Promotion Div of New Student Intake (HUMAS/PPMB) as the management will take further step in anticipating each year’s new student intake target. Keywords: Data Mining, DSS System, Support Vector Machine, ID3 Algorithm