Model Klasifikasi Berbasis Machine Learning Untuk Perpanjangan Masa Jabatan Kepala Sekolah Menggunakan Algoritma C4.5

Abstract

The extension assessment process of the tenure’s school principal at Dinas Pendidikan Kabupaten Musi Rawas runs manually, so it takes more than 1 (one) month to get the results. This study aims to model a classification based on Machine Learning technique using the C4.5 Decission Tree algorithm to make it easier for supervisors to provide recommendations whether the principal’s position has Extended or Not Extended status. The research design uses the CRISP-DM concept which is adopted to the needs of the research objectives. The resulting model has 15 rules which are used as the basis for forming a Decision Tree. Model validation measurement was tested using the Confusion Matrix method and it provides an accuracy of 83,3%.