MENENTUKAN PILIHAN SEKOLAH DIDALAM PENERIMAAN PESERTA DIDIK BARU DENGAN MENGGUNAKAN METODE NAÏVE BAYES DAN K-NEAREST NEIGHBOR (Studi Kasus: PPDB Online Jenjang SMP Kota Metro)

Abstract

Acceptance of New Learners (PPDB) is an important process to attract new students according to the criteria and characteristics of the school. PPDB needs to be implemented in an objective, transparent, accountable and non-discriminatory because it concerns the public interest, that is education. Following the current technological advances, PPDB has been held online. With the online PPDB is expected to provide solutions to parents of students for the realization of PPDB is transparent, accountable and not discriminatory. By using the online PPDB system PPDB results can be seen by anyone and anywhere. In addition, with the online PPDB will provide convenience in the registration process and the selection process because registration and selection process can be done and viewed anywhere so prospective students or parents will be easier to register.

This study aims to provide recommendations to parents in the implementation of PPDB junior high school in order to facilitate parents in determining the choice of schools in accordance with the circumstances of prospective students. In this study, the authors use the Naïve Bayes method and K-Nearest Neighbor which is the method used when determining the selection of schools in the process of acceptance of online PPDB. Naive Bayes will provide the school choice recommendation with the greatest probability of the calculation whereas K-NN sees the closeness of the data to the data of trying and the closest data to be used to provide recommendations to prospective students or parents / guardians. The steps of determining the probability of Bayes is to calculate the probability of a certain domicile, the origin of a particular school, again, the achievement of the data of student tryouts received at each junior high school SMP Negeri 1- SMP Negeri 10 which then the highest probability will be taken as a recommendation. While in K-NN use Euclidean distance equation to calculate distance data to data trying and take data closest to data trying as a recommendation.

Keywords: PPDB, K-NN, Naive Bayes