IMPLEMENTASI ALGORITMA DECISION TREE C4.5 UNTUK PREDIKSI KELULUSAN MAHASISWA: STUDI EKSPERIMENTAL

Timely graduation is an important indicator of higher education management success. This study aims to implement and evaluate the Decision Tree C4.5 algorithm for predicting student graduation based on academic data. The research employs a quantitative approach with an experimental method using a secondary dataset from Kaggle platform consisting of 51 student records. The variables include Grade Point Average (GPA), attendance percentage, final exam scores, academic activities, and organizational involvement. Implementation was conducted using Python with scikit-learn library. Results show that the Decision Tree C4.5 model achieves 96% accuracy with GPA as the most dominant factor. This study demonstrates the feasibility of C4.5 algorithm for graduation prediction and can serve as a baseline for implementation with real institutional data.