Perbandingan Metode K-Nearest Neighbor dan Support Vector Machine Untuk Memprediksi Penerima Beasiswa Keringanan UKT

Abstract

Scholarships are financial assistance provided to individuals, pupils, or scholars to extend their education. These may be provided by government agencies or the colleges themselves to students. One component that ensures quality human resources is formal education. The purpose of scholarships is to help disadvantaged or underprivileged students. Scholarship providers usually give some consideration to the student's level of difficulty, such as parents' salary and number of siblings. Due to the large number of applications for relief scholarships and strict assessment criteria, not all students who apply can be accepted. Scholarship application selection officers often have difficulty determining which students are worthy of receiving a scholarship. While the quota for scholarship recipients for this study program is always limited, applications for student UKT relief scholarships continue to increase every semester. This application came from students with poor economic conditions. To select UKT relief scholarship application documents, you have to consider various criteria and use manual methods which are less effective and require more time to determine the results. This research aims to make a comparison between the K-Nearest Neighbor and Support Vector Machine classification algorithms in determining recipients of UKT relief scholarships for undergraduate students in the Legal Sciences Study Program, Faculty of Law, Sebelas Maret University using the RapidMiner application. The accuracy results obtained using the RapidMiner application that have been carried out, the K-NN method produces an accuracy of 92.92%, while the SVM method produces an accuracy of 85.84%, so the K-NN method is the best method in classification for predicting recipients of UKT relief scholarships for students in the program. Bachelor of Law studies.