Implementasi Light Gradient Boosting Machine Untuk Klasifikasi Penyakit Diabetes Melitus Dengan Deteksi Outlier Robust Random Cut Forest
DOI:
https://doi.org/10.14421/fourier.2026.152.92-102Keywords:
Diabetes Mellitus, Classification, LightGBM, RRCFAbstract
Diabetes melitus merupakan penyakit tidak menular yang ditandai dengan tingginya kadar glukosa dalam darah dan menjadi salah satu masalah kesehatan utama di Indonesia. Tingginya jumlah penderita menuntut adanya metode deteksi dini yang akurat untuk mendukung pencegahan komplikasi serta membantu pengambilan keputusan di bidang kesehatan. Penelitian ini bertujuan mengevaluasi efektivitas model Light Gradient Boosting Machine dalam mengklasifikasikan penyakit diabetes melitus. Sebelum proses klasifikasi, dilakukan deteksi dan penanganan outlier menggunakan Robust Random Cut Forest dengan persentase penanganan sebesar 5%, serta penyeimbangan distribusi kelas menggunakan Synthetic Minority Over-sampling Technique. Evaluasi model dilakukan menggunakan K-Fold Cross Validation sebanyak lima lipatan. Selain itu, dilakukan pengujian beberapa kombinasi parameter, yaitu learning rate sebesar 0.05, 0.075, dan 0.1, num leaves sebanyak 2, 4, 6, 8, 10, 12, 14, dan 16, serta min data in leaf sebesar 10, 20, 40, dan 80 untuk memperoleh model terbaik. Hasil penelitian menunjukkan bahwa model terbaik mampu menghasilkan akurasi sebesar 85.36%, sensitivitas sebesar 79.03%, dan spesifisitas .sebesar 8577%. Penanganan outlier dan penyeimbangan data meningkatkan kemampuan model dalam mengenali kelas minoritas tanpa menurunkan kinerja klasifikasi secara signifikan. Dengan demikian, Light Gradient Boosting Machine efektif digunakan sebagai metode klasifikasi diabetes melitus dan berpotensi mendukung deteksi dini penyakit secara lebih akurat.
Downloads
References
S. Supriyo, M. Y. Ibrahim, P. N. D. Santoso, and S. U. Dwiningsih, “Diabetes Self Management Education (Dsme) Upaya Preventif Untuk Remaja Terhindar Diabetes Melitus,” J. Lintas Keperawatan, vol. 5, no. 1, pp. 355–361, 2024, doi: 10.31983/jlk.v5i1.11403.
IDF, “Diabetes in Indonesia,” International Diabetes Federation, 2024. https://idf.org/our-network/regions-and-members/western-pacific/members/indonesia/ (accessed Sep. 01, 2025).
H. Yun, J. won Park, and J. K. Kim, “A Comparative Evaluation of HbA1c Measurement Methods and Their Implications for Diabetes Management,” Diagnostics, vol. 13, no. 22, 2023, doi: 10.3390/diagnostics13223449.
N. Cassim, L. M. Coetzee, and D. K. Glencross, “Categorising specimen referral delays for CD4 testing: How inter-laboratory distances and travel times impact turn-around time across a national laboratory service in South Africa,” Afr. J. Lab. Med., vol. 9, no. 1, pp. 1–7, 2021, doi: 10.4102/AJLM.V9I1.1120.
E. Alpaydin, “Introduction of Machine Learning,” in The MIT Press, 3rd ed.The MIT Press, 2014.
A. Ma and F. Fauzi, “Perbandingan Klasifikasi Random Forest , Support Vector Machines , dan LGBM Pada Klasifikasi Kualitas Udara di Jakarta Comparison of Random Forest , Support Vector Machines , and LGBM Classification for Air,” J. Sist. dan Teknol. Inf. Indones., vol. 9, no. 2, pp. 99–108, 2024, [Online]. Available: https://ejurnal.unmuhjember.ac.id/index.php/JUSTINDO/article/view/1912
A. Al Tawil, L. Almazaydeh, B. Alqudah, A. Z. Abualkishik, and A. A. Alwan, “Predictive modeling for breast cancer based on machine learning algorithms and features selection methods,” Int. J. Electr. Comput. Eng., vol. 14, no. 2, pp. 1937–1947, 2024, doi: 10.11591/ijece.v14i2.pp1937-1947.
X. Xiong et al., “Application of LightGBM hybrid model based on TPE algorithm optimization in sleep apnea detection,” Front. Neurosci., vol. 18, no. February, pp. 1–17, 2024, doi: 10.3389/fnins.2024.1324933.
S. Vashisth and A. Goyal, “Dynamic Anomaly Detection Using Robust Random Cut Forests in Resource-Constrained IoT Environments,” Inform., vol. 48, no. 23, pp. 107–120, 2024, doi: 10.31449/inf.v48i23.6862.
I. Fadillah, Nurahmi, and L. B. Kurniawan, “Correlation between Serum Visfatin Level with Glycemic Control in Type 2 Diabetes Mellitus Subjects,” Indones. J. Clin. Pathol. Med. Lab., vol. 31, no. 3, pp. 299–303, 2025, doi: 10.24293/ijcpml.v31i3.2304.
B. De la Cruz Concepcion, Y. A. F. Cortez, M. I. B. Bonilla, J. M. M. Bello, and M. E. Rojo, “Insulin: A connection between pancreatic cells and the hypothalamus,” World J. Diabetes, vol. 14, no. 2, pp. 76–91, 2023, doi: 10.4239/wjd.v14.i2.76.
K. Kesehatan, “Kariadi Podcast Episode 35- Diabetes Melitus,” 2024. https://kms.kemkes.go.id/pengetahuan/detail/66b2f0c6108c3c91f387d845 (accessed Oct. 19, 2025).
S. P. Collins et al., “BUKU SAKU DIABETES MELITUS UNTUK AWAM,” in UNS Press, 2021, pp. 167–186.
S. Guha, N. Mishra, and G. Roy, “Robust Random Cut Forest Based Anomaly Detection On Streams,” vol. 48, 2016.
H. Vardhan and J. Sztipanovits, “Reduced Robust Random Cut Forest for Out-of-Distribution Detection in Machine Learning Models,” Int. J. Mach. Learn. Comput., vol. 12, no. 5, 2022, doi: 10.18178/ijmlc.2022.12.5.1104.
Hartono, “Modul Digital Machine Learning,”
V. W. Lumumba, D. Kiprotich, N. G. Makena, and M. D. 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. 127–137, 2024.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Siti Nurus Shofiyah, Putroue Intan
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
