Perbandingan CatBoost dan Elastic Net untuk Estimasi Komposisi Tubuh Berbasis Antropometri
DOI:
https://doi.org/10.30865/json.v7i4.9887Keywords:
antropometri, CatBoost, Elastic Net, komposisi tubuh, machine learningAbstract
Pemantauan komposisi tubuh diperlukan untuk memberi gambaran kondisi tubuh yang lebih informatif dibandingkan BMI, terutama terkait persentase lemak dan massa otot. Pemeriksaan standar seperti Dual-Energy X-ray Absorptiometry (DXA) memiliki keterbatasan biaya dan aksesibilitas, sehingga diperlukan pendekatan estimasi berbasis antropometri yang lebih praktis. Penelitian ini bertujuan membandingkan performa CatBoost dan Elastic Net dalam mengestimasi komposisi tubuh menggunakan dataset National Health and Nutrition Examination Survey (NHANES) 2017–2018. Data yang digunakan berjumlah 3.551 data setelah proses cleaning, terdiri atas delapan fitur input antropometri dan dua target berbasis DXA. Metode penelitian mengikuti CRISP-DM, meliputi pemahaman data, persiapan data, pemodelan, evaluasi, dan deployment sederhana. Data dibagi menjadi 2.840 data latih dan 711 data uji, kemudian model dievaluasi menggunakan MAE, RMSE, dan R². Hasil pengujian menunjukkan CatBoost memperoleh performa terbaik dengan Mean MAE 1,588891, Mean RMSE 2,044905, dan Mean R² 0,862830, sedangkan Elastic Net memperoleh Mean MAE 1,888025, Mean RMSE 2,408882, dan Mean R² 0,819884. Selain itu, prototipe berbasis Streamlit dikembangkan untuk simulasi inferensi terhadap data antropometri baru. Hasil penelitian ini masih terbatas pada dataset NHANES 2017–2018 sehingga generalisasi ke populasi lain memerlukan validasi lanjutan
References
[1] A. M. González-Martin et al., “Anthropometric Measurements for Predicting Low Appendicular Lean Mass Index for the Diagnosis of Sarcopenia : A Machine Learning Model,” J. Funct. Morphol. Kinesiol., vol. 10, no. 3, pp. 1–22, 2025, doi: 10.3390/jfmk10030276.
[2] K. Yasmin, “Cardiometabolic Risk Factors in South Asians: An Epidemiological and Anthropological Study in an Urban Populace of Eastern India.,” medRxiv, 2024, doi: 10.1101/2024.12.06.24318633.
[3] L. Zeng, X. Guo, H. Wu, and C. Huang, “Machine learning-based estimation of trunk fat percentage and its association with cardiometabolic risk leveraging two large national cohorts,” Front. Nutr., vol. 13, no. January, pp. 1–10, 2026, doi: 10.3389/fnut.2026.1715570.
[4] N. Birk, B. Kulkarni, S. Bhogadi, A. Aggarwal, and G. K. Walia, “Machine learning-based equations for improved body composition estimation in Indian adults,” PLOS Digit. Heal., vol. 4, no. 6, pp. 1–14, 2025, doi: 10.1371/journal.pdig.0000671.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Jurnal Sistem Komputer dan Informatika (JSON)
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
This work is licensed under a Creative Commons Attribution 4.0 International License
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under Creative Commons Attribution 4.0 International License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (Refer to The Effect of Open Access).
