Comparison of C4.5, Artificial Neural Network, and Gaussian Naïve Bayes for Predicting Uncollectible Education Receivables with Grouped Permutation Importance Analysis

Authors

  • Sry Faslia Hamka Institut Teknologi dan Bisnis Muhammadiyah Wakatobi
  • amin irmawan institut teknologi dan bisnis muhammadiyah wakatobi

DOI:

https://doi.org/10.25077/TEKNOSI.v12i2.2026.312-326

Keywords:

piutang pendidikan;, C4.5, klasifikasi pembelajaran mesin;, Artficial Neural Network

Abstract

Pengelolaan piutang biaya pendidikan merupakan tantangan operasional bagi perguruan tinggi, khususnya dalam mengidentifikasi secara dini mahasiswa yang berpotensi mengalami ketidaklancaran pembayaran. Penelitian ini bertujuan membandingkan kemampuan prediktif algoritma C4.5, Artificial Neural Network (ANN), dan Gaussian Naïve Bayes dalam mengklasifikasikan status piutang biaya pendidikan mahasiswa, serta menganalisis kontribusi atribut prediktor secara konsisten lintas-model. Data terdiri atas 258 rekaman mahasiswa dari Institut Teknologi dan Bisnis Muhammadiyah Wakatobi yang mengintegrasikan karakteristik akademik, wilayah tempat tinggal, kondisi sosial ekonomi wali, serta jumlah dan umur piutang UKT dan BPP. Evaluasi menggunakan holdout terstratifikasi 80:20 dan 90:10 serta RepeatedStratifiedKFold 5×10 sebagai prosedur utama. Seluruh algoritma dilatih pada pembagian data yang identik dengan tahap imputasi, encoding, standardisasi, dan oversampling ditempatkan dalam pipeline pelatihan. Kinerja dievaluasi menggunakan accuracy, precision, recall, F1-score, balanced accuracy, dan PR-AUC. Analisis interpretabilitas menggunakan grouped permutation importance diterapkan seragam pada ketiga algoritma. ANN memperoleh kinerja tertinggi dengan accuracy 82,09% ± 6,17%, recall 82,71%, F1-score 78,62%, balanced accuracy 82,20%, dan PR-AUC 91,32%. C4.5 menghasilkan kinerja mendekati ANN dengan accuracy 81,04% ± 4,98% dan PR-AUC 89,08%, disertai simpangan baku lebih kecil yang mencerminkan prediksi lebih stabil. Gaussian Naïve Bayes menghasilkan recall 80,58% namun balanced accuracy hanya 53,86% dan PR-AUC 46,99%, mengindikasikan kecenderungan prediksi positif berlebihan. Jumlah Piutang UKT Mahasiswa merupakan atribut paling berpengaruh secara konsisten pada ketiga algoritma, diikuti Umur Piutang UKT dan Jumlah Piutang BPP. Atribut akademik, wilayah, dan sosial ekonomi memberikan kontribusi yang kecil atau tidak stabil. ANN dipilih sebagai model utama untuk kepentingan operasional yang memprioritaskan deteksi piutang tak tertagih, sedangkan C4.5 menjadi alternatif kompetitif apabila institusi mengutamakan keterpahaman aturan klasifikasi. Hasil penelitian memberikan dasar ilmiah bagi pengembangan sistem pendukung keputusan prioritas pemeriksaan piutang dengan tetap mensyaratkan verifikasi manual sebelum tindak lanjut diambil.

References

P. A. Prayesy and E. S. Negara, “Classification of the Fluency Multipurpose of Bank Mandiri Credit Payments Based on Debtor Preferences Using Naive Bayes and Neural Network Method,” vol. 7, no. 1, pp. 7–16, Jun. 2022, doi: 10.15575/join.v7i1.762.

I. T. A. Nur, N. Y. Setiawan, and F. A. Bachtiar, “Perbandingan Performa Metode Klasifikasi SVM, Neural Network, dan Naive Bayes untuk Mendeteksi Kualitas Pengajuan Kredit di Koperasi Simpan Pinjam,” J. Teknol. Inf. Dan Ilmu Komput., vol. 6, no. 4, p. 444, Jul. 2019, doi: 10.25126/jtiik.2019641352.

X. Zhu, Q. Chu, X. Song, P. Hu, and L. Peng, “Explainable prediction of loan default based on machine learning models,” Data Sci. Manag., vol. 6, no. 3, pp. 123–133, Sep. 2023, doi: 10.1016/j.dsm.2023.04.003.

X. Zhang et al., “Data-Driven Loan Default Prediction: A Machine Learning Approach for Enhancing Business Process Management,” Systems, vol. 13, no. 7, p. 581, Jul. 2025, doi: 10.3390/systems13070581.

V. S. Ginting, K. Kusrini, and E. T. Luthfi, “PENERAPAN ALGORITMA C4.5 DALAM MEMPREDIKSI KETERLAMBATAN PEMBAYARAN UANG SEKOLAH MENGGUNAKAN PYTHON,” J. Teknol. Inf., vol. 4, no. 1, pp. 1–6, Jun. 2020, doi: 10.36294/jurti.v4i1.1101.

J. R. Quinlan, C4.5: Programs for Machine Learning. San Mateo, CA: Morgan Kaufmann Publishers, 1993.

J. R. Quinlan, “Induction of Decision Trees,” Mach. Learn., vol. 1, no. 1, pp. 81–106, 1986, doi: 10.1007/BF00116251.

G. E. Hinton, “Connectionist Learning Procedures,” Artif. Intell., vol. 40, no. 1–3, pp. 185–234, 1989, doi: 10.1016/0004-3702(89)90049-0.

scikit-learn developers, “1.17. Neural Network Models (Supervised) — MLPClassifier.” 2024. Accessed: Jun. 21, 2026. [Online]. Available: https://scikit-learn.org/stable/modules/neural_networks_supervised.html

F. Pedregosa et al., “Scikit-learn: Machine Learning in Python,” J. Mach. Learn. Res., vol. 12, pp. 2825–2830, 2011, doi: 10.5555/1953048.2078195.

A. Rahmayanti, L. Rusdiana, and S. Suratno, “PERBANDINGAN METODE ALGORITMA C4.5 DAN NAÏVE BAYES UNTUK MEMPREDIKSI KELULUSAN MAHASISWA,” Walisongo J. Inf. Technol., vol. 4, no. 1, pp. 11–22, Aug. 2022, doi: 10.21580/wjit.2022.4.1.9654.

L. Breiman, “Random Forests,” Mach. Learn., vol. 45, no. 1, pp. 5–32, 2001, doi: 10.1023/A:1010933404324.

B. Gregorutti, B. Michel, and P. Saint-Pierre, “Grouped Variable Importance with Random Forests and Application to Multiple Functional Data Analysis,” Comput. Stat. Data Anal., vol. 90, pp. 15–35, 2015, doi: 10.1016/j.csda.2015.04.002.

J. Davis and M. Goadrich, “The Relationship Between Precision-Recall and ROC Curves,” in Proceedings of the 23rd International Conference on Machine Learning (ICML ’06), New York, NY, USA: ACM, 2006, pp. 233–240. doi: 10.1145/1143844.1143874.

M. Imani, M. Joudaki, A. Bagheri, and H. R. Arabnia, “Why ROC-AUC Is Misleading for Highly Imbalanced Data: In-Depth Evaluation of MCC, F2-Score, H-Measure, and AUC-Based Metrics Across Diverse Classifiers,” Technologies, vol. 14, no. 1, p. 54, Jan. 2026, doi: 10.3390/technologies14010054.

S. Szeghalmy and A. Fazekas, “A Comparative Study of the Use of Stratified Cross-Validation and Distribution-Balanced Stratified Cross-Validation in Imbalanced Learning,” Sensors, vol. 23, no. 4, p. 2333, Feb. 2023, doi: 10.3390/s23042333.

V. Lumumba, D. Kiprotich, M. Mpaine, N. Makena, and M. Kavita, “Comparative Analysis of Cross-Validation Techniques: LOOCV, K-folds Cross-Validation, and Repeated K-folds Cross-Validation in Machine Learning Models,” Am. J. Theor. Appl. Stat., vol. 13, no. 5, pp.

–137, Oct. 2024, doi: 10.11648/j.ajtas.20241305.13.

G. Lemaître, F. Nogueira, and C. K. Aridas, “Imbalanced-learn: A Python Toolbox to Tackle the Curse of Imbalanced Datasets in Machine Learning,” J. Mach. Learn. Res., vol. 18, no. 17, pp. 1–5, 2017.

M. Rosenblatt, L. Tejavibulya, R. Jiang, S. Noble, and D. Scheinost, “Data leakage inflates prediction performance in connectome-based machine learning models,” Nat. Commun., vol. 15, no. 1, p. 1829, Feb. 2024, doi: 10.1038/s41467-024-46150-w.

S. Kapoor and A. Narayanan, “Leakage and the reproducibility crisis in machine-learning-based science,” Patterns, vol. 4, no. 9, p. 100804, Sep. 2023, doi: 10.1016/j.patter.2023.100804.

F. R. Pratikto, “Oversampling Sintetis Berbasis Kopula untuk Model Klasifikasi dengan Data yang Tidak Seimbang,” J. Rekayasa Sist. Ind., vol. 12, no. 1, pp. 1–10, Apr. 2023, doi: 10.26593/jrsi.v12i1.6380.1-10.

K. K. Nicodemus, J. D. Malley, C. Strobl, and A. Ziegler, “The Behaviour of Random Forest Permutation-Based Variable Importance Measures Under Predictor Correlation,” BMC Bioinformatics, vol. 11, p. 110, 2010, doi: 10.1186/1471-2105-11-110.

S. Mestiri, “Credit scoring using machine learning and deep Learning-Based models,” Data Sci. Finance Econ., vol. 4, no. 2, pp. 236–248, 2024, doi: 10.3934/DSFE.2024009.

T. T. Do, G. Babaei, and P. Pagnottoni, “Explainable Machine Learning for Credit Risk Management When Features are Dependent,” Meas. Interdiscip. Res. Perspect., vol. 22, no. 4, pp. 315–340, Oct. 2024, doi: 10.1080/15366367.2023.2261186.

C. Molnar, G. König, B. Bischl, and G. Casalicchio, “Model-agnostic feature importance and effects with dependent features: a conditional subgroup approach,” Data Min. Knowl. Discov., vol. 38, no. 5, pp. 2903–2941, Sep. 2024, doi: 10.1007/s10618-022-00901-9.

I. M. Zubair, Y.-S. Lee, and B. Kim, “A New Permutation-Based Method for Ranking and Selecting Group Features in Multiclass Classification,” Appl. Sci., vol. 14, no. 8, p. 3156, Apr. 2024, doi: 10.3390/app14083156.

Submitted

2026-06-27

Accepted

2026-08-20

Published

2026-08-28

How to Cite

[1]
S. Faslia Hamka and amin irmawan, “Comparison of C4.5, Artificial Neural Network, and Gaussian Naïve Bayes for Predicting Uncollectible Education Receivables with Grouped Permutation Importance Analysis”, TEKNOSI, vol. 12, no. 2, pp. 312–326, Aug. 2026.

Similar Articles

1 2 3 4 5 6 7 8 9 10 > >> 

You may also start an advanced similarity search for this article.