Comparison of K-Nearest Neighbor Classification Methods and Support Vector Machine in Predicting Students’ Study Period.

Abstract

State Universities and Private Universities compete fiercely to produce quality students in line with the development of the world of education in Indonesia. Universities strive to improve quality and provide the best education to students and the number of students who graduate on time or not. In this research, a comparative test of the performance of the accuracy values of the K-Nearest Neighbor algorithm and the Support Vector Machine was carried out as a classification method for predicting the study period of students in the Bachelor of Law study program, Faculty of Law. Law, Sebelas Maret University, Surakarta, Indonesia using the RapidMiner application. In this study, a comparison of two classification methods was used, namely K-Nearest Neighbor and Support Vector Machine with 433 student data used. The data is divided into 70% training data and 30% test data. The test results for the highest K-NN prediction accuracy value were at K=5, namely 98.45%. While for the Support Vector Machine method, the accuracy value using the SVM model was 96.90%. Therefore, the results of this research are included in the good category in producing high accuracy, so that the contribution of the K-NN modeling research results using the value K=5 is getting the best accuracy compared to the SVM method using the SVM in predicting student study periods. class of 2021, Bachelor of Law study program, Faculty of Law, Sebelas Maret University.

Keywords

• classification
• student study period
• svm
• knn