<![CDATA[Penelitian ini bertujuan untuk menerapkan teknik data mining dalam memprediksi jumlah pendaftar Penerimaan Peserta Didik Baru (PPDB) di SMKN3 Metro menggunakan algoritma machine learning, khususnya Decision Tree (C4.5). Masalah utama yang dihadapi adalah tantangan pengelolaan data historis dan keterbatasan kapasitas sekolah dalam merencanakan penerimaan siswa secara efektif. Metode penelitian meliputi pengumpulan data historis pendaftaran, pra-pemrosesan data, penerapan algoritma machine learning, serta evaluasi kinerja model menggunakan metrik seperti Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), dan koefisien determinasi (R²).Hasil penelitian menunjukkan bahwa model Decision Tree (C4.5) memiliki performa terbaik dibandingkan algoritma lain, dengan nilai MSE sebesar 290,948, RMSE 17,057, MAE 11,096, dan R² sebesar 0,893. Akurasi prediksi yang tinggi ini menunjukkan potensi besar algoritma tersebut dalam mendukung pengelolaan PPDB secara lebih efisien. Penelitian ini diharapkan dapat menjadi solusi inovatif bagi SMKN3 Metro dalam merencanakan penerimaan siswa baru dan optimalisasi sumber daya sekolah. Selain itu, model ini dapat menjadi referensi bagi institusi pendidikan lain dalam mengadopsi teknologi serupa.Kata Kunci: Data Mining, Prediksi Pendaftar, PPDB, Decision Tree, SMKN3 MetroAbstractThis research focuses on implementing data mining to predict the number of registrants for new student admissions (PPDB) at SMKN3 Metro using the C4.5 machine learning algorithm. The study aims to address annual challenges in data management and school capacity limitations. By leveraging historical registration data, an accurate predictive model is developed to assist the school in planning student admissions more effectively. The methodology includes data collection and preprocessing, application of the C4.5 algorithm, and model performance evaluation based on prediction accuracy. Preliminary results indicate that the C4.5 algorithm outperforms other models, achieving a Mean Squared Error (MSE) of 290.948, Root Mean Squared Error (RMSE) of 17.057, and a coefficient of determination (R²) of 0.893. These findings demonstrate the model's reliability in estimating the number of registrants for key competencies such as Software Engineering and Computer Network Engineering. This implementation is expected to improve the efficiency of the PPDB process and resource planning at SMKN3 Metro, while providing a practical application of data mining and machine learning in educational management.Keywords: Data Mining, PPDB Prediction, Machine Learning, C4.5 Algorithm, SMKN3 Metro]]>
