Implementasi Algoritma Naive Bayes Untuk Klasifikasi Penerima Beasiswa (Studi Kasus Universitas Hamzanwadi)

Abstract

Every year the University offers various types of scholarships to its students, including Hamzanwadi Selong University, East Lombok. One type of scholarship offered is the Bidikmisi scholarship (KIP/K) which is intended for students who have middle to lower economic levels but have good academic achievement or potential. Every year the number of applicants for this scholarship continues to increase, but the number received each year is limited. The large number of piles of files applying for scholarships and the manual selection process tends to be ineffective and efficient and the results of the selection are inaccurate, so a system is needed that is able to assist in the selection process quickly, easily and on target. The method used is CRIS-DM with the Naive Bayes Classifier algorithm modeling, this method is an approach that refers to the Bayes theorem which combines previous knowledge with new knowledge. The variables consist of 7 attributes, namely: Name, DTKS status, achievement, parents' occupation, total income of parents, home ownership, and number of family dependents[1]. Testing was carried out using k-fold cross validation, and the highest accuracy results were obtained from k-fold 4 of 91.43%, while the AUC was 0.996% with a very good diagnostic classification. Thus it can be interpreted that the Naive Bayes algorithm is very well used in the selection of scholarships for bidikmisi scholarships at Hamzanwadi University