Prediksi Risiko Keterlambatan Kelulusan Mahasiswa S1 Pendidikan Teknik Bangunan Di UNJ Menggunakan Machine Learning Berbasis Data SIAKAD

OCTAVIANUS BONGGO ADHISUKMA, . (2026) Prediksi Risiko Keterlambatan Kelulusan Mahasiswa S1 Pendidikan Teknik Bangunan Di UNJ Menggunakan Machine Learning Berbasis Data SIAKAD. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text COVER.pdf Download (1MB)
	Text BAB I.pdf Download (227kB)
	Text BAB II.pdf Restricted to Registered users only Download (590kB) | Request a copy
	Text BAB III.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text BAB IV.pdf Restricted to Registered users only Download (786kB) | Request a copy
	Text BAB V.pdf Restricted to Registered users only Download (249kB) | Request a copy
	Text DAFTAR PUSTAKA.pdf Download (245kB)
	Text LAMPIRAN.pdf Restricted to Registered users only Download (1MB) | Request a copy

Abstract

Penelitian ini bertujuan untuk mengembangkan dan mengevaluasi model prediksi risiko keterlambatan kelulusan mahasiswa S1 Pendidikan Teknik Bangunan di Universitas Negeri Jakarta menggunakan metode machine learning berbasis data Sistem Informasi Akademik (SIAKAD). Tingginya angka keterlambatan kelulusan, yang mencapai 186 mahasiswa dari total 276 pada data yang diteliti, menggarisbawahi urgensi penanganan isu ini. Metode penelitian ini melibatkan pengumpulan dan pemrosesan data akademik mahasiswa angkatan 2017-2021, mencakup Indeks Prestasi Semester (IPS), jumlah SKS, data mata kuliah, dan biaya kuliah, untuk kemudian dilakukan rekayasa fitur (feature engineering). Lima algoritma machine learning dievaluasi: Random Forest, Extra Trees, Gradient Boosting, XGBoost, dan CatBoost. Hasil evaluasi menunjukkan bahwa algoritma Random Forest menawarkan performa terbaik dengan keseimbangan antara akurasi, presisi, recall, dan kestabilan model, yang dibuktikan melalui analisis feature importance di mana IPS semester 2, 3, dan 4 beserta perubahannya (Delta IPS) menjadi faktor prediktif dominan. Model Random Forest ini diharapkan dapat berfungsi sebagai early warning system bagi program studi untuk mengidentifikasi mahasiswa berisiko terlambat lulus sejak dini, sehingga intervensi akademik yang tepat sasaran dapat dilakukan guna meningkatkan efisiensi dan kualitas lulusan. ***** This study aims to develop and evaluate a machine learning-based model for predicting the risk of delayed graduation among undergraduate students in the Building Engineering Education Program at Universitas Negeri Jakarta using data from the Academic Information System (SIAKAD). The high rate of delayed graduation, with 186 out of 276 students in the analyzed dataset experiencing delayed graduation, highlights the urgency of addressing this issue. The research methodology involved collecting and preprocessing academic data from students admitted between 2017 and 2021, including Semester Grade Point Average (GPA), accumulated credit units (SKS), course records, and tuition fee data, followed by a feature engineering process. Five machine learning algorithms were evaluated, namely Random Forest, Extra Trees, Gradient Boosting, XGBoost, and CatBoost. The evaluation results indicate that the Random Forest algorithm achieved the best overall performance by providing a balanced combination of accuracy, precision, recall, and model stability. Feature importance analysis further revealed that the Semester Grade Point Average (GPA) in the second, third, and fourth semesters, along with the changes in GPA (Delta GPA), were the most influential predictive factors. The proposed Random Forest model is expected to serve as an early warning system for the study program to identify students at risk of delayed graduation at an early stage, enabling timely academic interventions to improve graduation efficiency and the overall quality of graduates.

Item Type:	Thesis (Sarjana)
Additional Information:	1). Murien Nugraheni, S.T., M.Cs. ; 2). Dr. M. Agphin Ramadhan, M.Pd.
Subjects:	Pendidikan > Evaluasi Pendidikan Pendidikan > Teori, Penelitian Pendidikan Teknologi dan Ilmu Terapan > Teknik Komputer
Divisions:	FT > S1 Pendidikan Teknik Bangunan
Depositing User:	Octavianus Bonggo Adhisukma .
Date Deposited:	26 Aug 2026 01:40
Last Modified:	26 Aug 2026 01:40
URI:	http://repository.unj.ac.id/id/eprint/73113

Actions (login required)

View Item