The Support Vector Machine (SVM) And Random Forest Methods For Classification Graduation Rate

Abstract

Efforts towards an independent nation with high competitiveness can’t be separated from educational programs. Therefore, education must be able to produce quality graduates who have knowledge, master technology, and have technical skills, and adequate life skills. The timeliness of students in completing their studies is one of the supports for assessing the quality of higher education. Classification analysis can be used to predict whether a student is said to pass on time or not. Support Vector Machine (SVM) and Random Forest methods are part of the classification method. SVM and Random Forest classification analysis is done by using historical data alumni from FMIPA UNM of the graduation year 2019-2020 which come from the Administration, Academic and Student Affair Bureau of UNM. SVM accuracy level of RBF kernel with optimum value C = 1 and gamma = 1 is 68% and Random Forest accuracy with optimum value m = 2 and k = 500 is 72%. Therefore, the best method for determining the accuracy of the study duration of FMIPA UNM students is Random Forest..