IMPLEMENTASI ALGORITMA K-NEAREST NEIGHBOR DAN NAIVE BAYES DALAM MEMPREDIKSI STATUS SELEKSI PADA PPDB

Abstract

Abstracts, This study aims to implement the K-Nearest Neighbor (KNN) and Naive Bayes algorithms to predict the selection status of New Student Admissions (PPDB) at the junior high school level in Cianjur Regency. PPDB is an annual agenda that plays a crucial role in determining the transition of students to higher education levels. However, the selection process often poses challenges, particularly due to limited information and subjectivity in decision-making by students and parents. This research adopts a quantitative approach by utilizing historical registration data from 20 public junior high schools in Cianjur Regency. The research procedure includes data collection, preprocessing, implementation of the KNN and Naive Bayes algorithms, and evaluation using the Confusion Matrix. The results indicate that both algorithms are capable of predicting students’ acceptance status through the zoning and achievement tracks with accuracy levels above 85%. Naive Bayes demonstrates advantages in computational efficiency, while KNN provides greater flexibility in handling variations in data. The developed prediction system is expected to assist students and parents in determining the most suitable school objectively and support schools and education authorities in providing data-driven recommendations. Furthermore, this study reinforces findings from previous research, emphasizing the potential of data mining as an effective approach to support educational selection processes and decision-making.