Metode Data Mining untuk Seleksi Calon Mahasiswa Baru pada Penerimaan Mahasiswa Baru di Universitas Muhammadiyah Gresik

Abstract

Muhammadiyah University of Gresik is one of the educational institutions in Gresik. Every year, the ratio of new students to graduates is not the same, which can affect the accreditation of the campus. To address this issue, a prediction is made on the data of prospective new students to detect whether they can graduate on time or not. A comparison of the classification results is performed using the K-Nearest Neighbor and Naive Bayes methods. From the implementation and testing, Naive Bayes achieves an accuracy of 72%, while the K- Nearest Neighbor method achieves an accuracy of 64%. Therefore, Naive Bayes is better at classifying the data of prospective new students compared to K-Nearest Neighbor.