PREDIKSI KELULUSAN JALUR MASUK PERGURUAN TINGGI BANDA ACEH (STUDY KASUS MAHASISWA BARU TAHUN AJARAN 2019 ) MENGGUNAKAN ALGORITMA SUPPORT VECTOR MACHINE (SVM)

Abstract

Nowadays there are several types of entry points for new student admissions at tertiary institutions. There are many ways that every prospective student can prepare to pass the PTN entrance selection exam. Therefore the research aim to predict the graduation of tertiary entrance by classification method using the SVM algorithm (Support Vector Machine) which is assisted by WEKA machine learning, using data of new students in the 2019 school year. The final results in this study there are two variables that have the most relationship both the PTN entry selection tutoring variable and the interest pathway variable, with the PTN entry selection tutoring variable having a Pearson correlation value of -0.180 ** and a significance value of 0.002, the interest pathway having an accuracy value of 0.311 ** and a significance value of 0,000. Then based on the results of cross-validation testing and percentage split SVM algorithm has very good accuracy with an average accuracy of 99% with an AUC (Area Under Curve) value of 0.9907