Predicting High School Student Interest in Private Universities Using Markov Chains and Data Mining

Hevlie Winda Nazry S; Muhammad Haris; Fatma Sari Hutagalung

doi:10.51574/jrip.v6i2.5696

Authors

Hevlie Winda Nazry S Data Science Study Program, Faculty of Computer Science and Information Technology, Universitas Muhammadiyah Sumatera Utara, Medan, Indonesia https://orcid.org/0009-0002-2584-8094
Muhammad Haris Information Technology Study Program, Faculty of Computer Science and Information Technology, Universitas Muhammadiyah Sumatera Utara, Medan, Indonesia https://orcid.org/0009-0004-5838-2344
Fatma Sari Hutagalung Information Technology Study Program, Faculty of Computer Science and Information Technology, Universitas Muhammadiyah Sumatera Utara, Medan, Indonesia

Keywords:

Educational Data Mining, Markov Chain, Student Enrollment, Private University, Transition Probability

Abstract

This study examines senior high school (SMA) students interest in attending private universities in Medan using an Educational Data Mining (EDM) approach. A first-order Markov chain model integrated with data mining techniques maps transitions among five states (Exploration, Consideration, Application, Re-enrollment, and Non-PTS migration) of student preferences. A stratified random sample of 2,850 SMA students was surveyed with a 20-item Likert questionnaire, and secondary historical enrollment data were used. K-Means clustering and decision tree algorithms were applied for data preprocessing and feature extraction. The Markov model achieved 89.47% accuracy, 88.20% precision, 90.15% recall, and an RMSE of 0.0824. The steady-state projection indicates 52.09% of students eventually enroll in private universities (PTS) while 47.79% go to public universities (PTN) or other non-PTS pathways; the remaining ~0.12% probability is distributed across the other three states, confirming that the 52.09% and 47.79% outcomes are dominant. These results have practical implications for private university policymakers in formulating data-driven recruitment strategies and mitigating the “leaky pipeline” of applicants.

Downloads

Download data is not yet available.

References

Baker, R. S. (2021). Educational Data Mining: Applications and trends in higher education. Springer.

Gandy, R., Harrison, P., & Goldring, H. (2019). Using Markov chains to analyze student transition and leaky pipeline dynamics. Higher Education Policy, 32(3), 321-340.

Han, J., Kamber, M., & Pei, J. (2022). Data mining: Concepts and techniques (4th ed.). Morgan Kaufmann.

Kitsantas, A., Cleary, T. J., & Parker, M. (2021). Predictive analytics in tertiary education decision-making. Journal of Educational Psychology, 113(4), 780-795.

LLDIKTI Region I. (2023). Laporan Kinerja Perekrutan dan Partisipasi Mahasiswa Baru Perguruan Tinggi Swasta Sumatera Utara. Lembaga Layanan Pendidikan Tinggi Wilayah I.

Montgomery, D. C., Runger, G. C., & Hubele, N. F. (2020). Engineering statistics and stochastic modeling (7th ed.). John Wiley & Sons.

Nadhiroh, N., Santoso, B., & Supriyanto, A. (2024). Institutional selection trends and shift dynamics in Indonesian private higher education. Jurnal Pendidikan Tinggi, 12(1), 88-102.

Nitisastro, M., & Wibowo, A. (2022). Educational decision trees for student preference classification. Indonesian Journal of Computer Science, 19(2), 112-125.

Pratama, A., & Suryani, E. (2023). Simulation modeling for private university capacity planning. Jurnal Teknologi Informasi dan Ilmu Komputer, 10(4), 789-798.

Rahmadani, F., & Sitorus, T. (2022). Survey of student decision factors in North Sumatra higher education institutions. Jurnal Manajemen Pendidikan, 11(2), 145-158.

Romero, C., & Ventura, S. (2020). Educational data mining and learning analytics: An updated survey. WIREs Data Mining and Knowledge Discovery, 10(3), e1355.

Setiawan, D., & Haryanto, E. (2021). Analysis of private university choice factors among high school graduates in Sumatra. Jurnal Evaluasi Pendidikan, 9(3), 201-215.

Siraj, F., Abdullah, S., & Ahmad, N. (2021). Markov chain modeling for prospective student enrollment and retention. International Journal of Information and Education Technology, 11(8), 375-382.

Sugiyono. (2022). Metode Penelitian Kuantitatif, Kualitatif, dan R&D. Alfabeta.

Tofa, M. (2022). Application of Naïve Bayes and K-Nearest Neighbors for predicting high school student progression to tertiary education. Journal of Educational Data Mining, 14(2), 45-58.

Wibisono, S., & Pratama, R. (2023). Market share forecasting of private higher education institutions using Markov chains. Jurnal Sistem Informasi, 15(1), 34-46.

Wijaya, C., & Kusuma, A. (2020). Student decision-making trajectory in higher education recruitment. Journal of Higher Education Organization and Management, 28(2), 140-155.

Zhao, L., & Otteson, K. (2024). Stochastic modeling of student enrollment trajectories in competitive academic markets. Computers & Education, 195, 104712.

Zhou, Y., & Chen, X. (2021). Cluster-driven Markov chains for predictive analytics in academic enrollment. IEEE Transactions on Learning Technologies, 14(5), 620-632.