District Level Child Nutrition Vulnerability in North Sumatra Using Composite Indeks, Machine Learning Benchmarking

Fanny Ramadhani; Sri Dewi; Dian Septiana; Diah Retno Wahyuningrum; Andy Satria

doi:10.30865/json.v7i4.9785

Authors

Fanny Ramadhani Universitas Negeri Medan
Sri Dewi Universitas Negeri Medan
Dian Septiana Universitas Negeri Medan
Diah Retno Wahyuningrum Universitas Negeri Medan
Andy Satria Universitas Dharmawangsa

DOI:

https://doi.org/10.30865/json.v7i4.9785

Keywords:

Child Nutrition, Composite Indek, Risk Mapping, Sensitivity Analysis, Machine Learning

Abstract

Child nutrition vulnerability is a multidimensional issue influenced by health outcomes, socioeconomic conditions, environmental factors, maternal education, dietary patterns, and access to health services. This study developed the Child Nutrition Vulnerability Index (IKGA) to map district level vulnerability in 33 districts or cities of North Sumatra during 2021-2023 and to examine the stability of the resulting priority ranking. Birth, severe malnutrition, and low birth weight data were obtained from the North Sumatra Provincial Health Office; sanitation was derived from Riskesdas; and supporting socioeconomic and service access variables were compiled from district level research and statistical datasets. The method included indicator transformation, annual min-max normalization, weighted composite aggregation, tertile-based relative risk classification, spatial visualization, one-at-a-time and Monte Carlo sensitivity analyses, and exploratory machine learning benchmarking using Random Forest, Decision Tree, and K-Means. The 2023 average IKGA was 0.3486, with the highest score in Nias (0.5573) and the lowest in Kota Binjai (0.1934). Monte Carlo sensitivity analysis produced an average Spearman rank correlation of 0.9957, indicating stable rankings under moderate weight variation. Random Forest reproduced the IKGA categories better than Decision Tree and K-Means, with 0.727 accuracy and 0.724 macro-F1. The IKGA provides an interpretable district level prioritization tool for SI-GIZI SIGAP, but the categories should be interpreted as relative provincial priorities rather than absolute nutritional risk thresholds.

References

[1] “Victora, C.G., Adair, L., Fall, C., Hallal, P.C., Martorell, R., Richter, L.M., et al. (2008) Maternal and Child Undernutrition Consequences for Adult Health and Human Capital. The Lancet, 371, 340-357. - References - Scientific Research Publishing.” Accessed: May 08, 2026. [Online]. Available: https://www.scirp.org/reference/referencespapers?referenceid=2762043

[2] WHO, “World Health Organization Growth Standards,” Acta Paediatr., pp. 5–6, 2006, Accessed: May 08, 2026. [Online]. Available: http://www.who.int/childgrowth/standards/Technical_report.pdf?ua=1

[3] “Levels and trends in child malnutrition: UNICEF/WHO/World Bank Group joint child malnutrition estimates: key findings of the 2023 edition.” Accessed: May 08, 2026. [Online]. Available: https://www.who.int/publications/i/item/9789240073791

[4] V. J. B. Matrins et al., “Long-Lasting Effects of Undernutrition,” Int. J. Environ. Res. Public Health, vol. 8, no. 6, p. 1817, 2011, doi: 10.3390/IJERPH8061817.

[5] X. Wang, B. Höjer, S. Guo, S. Luo, W. Zhou, and Y. Wang, “Stunting and ‘overweight’ in the WHO Child Growth Standards -malnutrition among children in a poor area of China,” Public Health Nutr., vol. 12, no. 11, pp. 1991–1998, Feb. 2009, doi: 10.1017/S1368980009990796.

[6] T. Beal, A. Tumilowicz, A. Sutrisna, D. Izwardy, and L. M. Neufeld, “A review of child stunting determinants in Indonesia,” Matern. Child Nutr., vol. 14, no. 4, p. e12617, Oct. 2018, doi: 10.1111/MCN.12617.

[7] F. Ramadhani, D. Septiana, S. N. Amalia, P. M. Fadilah, and A. Satria, “Spatial Clustering Analysis of Stunting in North Sumatra Based on Environmental Factors Using K-Means Algorithm,” Data Science: Journal of Computing and Applied Informatics, vol. 9, no. 2, pp. 18–25, Jul. 2025, doi: 10.32734/JOCAI.V9.I2-17179.

[8] F. Ramadhani, S. I. Al-Idrus, D. Septiana, Arnita, D. R. Wahyuningrum, and Salamah, “Machine Learning-Based Risk Prediction and Spatial Mapping of Stunting in North Sumatra Using a Strategic Policy Approach,” pp. 1–6, Jan. 2026, doi: 10.1109/ICCAI65301.2025.11278946.

[9] R. E. Caraka et al., “Bayesian Network analysis of spatial disparities in child nutrition status and health determinants across Indonesia,” International Journal of Applied Earth Observation and Geoinformation, vol. 144, p. 104936, Nov. 2025, doi: 10.1016/J.JAG.2025.104936.

[10] A. R. Nasution, A. Mardhiyah, and M. Rinaldi, “Exploration of development inequality models in north Sumatera,” 2025. [Online]. Available: www.ijafibs.pelnus.ac.id

[11] K. A. Hutasuhut, L. M. Siahaan, and J. Hutasuhut, “Economic Convergence Among Urban Areas In North Sumatera,” Journal of Sustainable Economics, vol. 3, no. 1, pp. 27–41, May 2025, doi: 10.32734/jse.v3i1.20286.

[12] V. Grigoriadis et al., “Developing the Sus-Health Index: a combined measure for describing environmental impact and nutritive value of foods and meals,” Philosophical Transactions of the Royal Society B: Biological Sciences, vol. 380, no. 1935, p. 20240160, Sep. 2025, doi: 10.1098/RSTB.2024.0160.

[13] J. M. Bôto, B. Neto, V. Miguéis, and A. Rocha, “Development of the Dietary Pattern Sustainability Index (DIPASI): A novel multidimensional approach for assessing the sustainability of an individual’s diet,” Sustain. Prod. Consum., vol. 50, pp. 139–154, Oct. 2024, doi: 10.1016/J.SPC.2024.07.029.

[14] C. Wehbe and H. Baroud, “Limitations and considerations of using composite indicators to measure vulnerability to natural hazards,” Scientific Reports 2024 14:1, vol. 14, no. 1, pp. 19333-, Aug. 2024, doi: 10.1038/s41598-024-68060-z.

[15] M. Wanyonyi, Z. N. Morris, F. M. Musyoka, and D. M. Kitavi, “Enhanced machine learning and hybrid ensemble approaches for Coronary Heart Disease prediction,” PLoS One, vol. 20, no. 12, p. e0328338, Dec. 2025, doi: 10.1371/JOURNAL.PONE.0328338.

[16] A. Podtschaske, B. Ulm, and S. Kagerbauer, “Pre-operative risk stratification from admission data: Clustering and KNN-modelling for identifying high risk patients before surgery,” J. Crit. Care, vol. 92, p. 155376, Apr. 2026, doi: 10.1016/J.JCRC.2025.155376.

[17] M. Al Mahmud, S. Chowdhury, A. A. M. Hamim, and A. Haque, “Prevalence and determinants of low birth weight and its association with child malnutrition in Bangladesh,” BMC Nutr., vol. 12, no. 1, p. 76, Dec. 2026, doi: 10.1186/S40795-026-01291-7.

[18] K. Cabello-Solorzano, I. Ortigosa de Araujo, M. Peña, L. Correia, and A. J. Tallón-Ballesteros, “The Impact of Data Normalization on the Accuracy of Machine Learning Algorithms: A Comparative Analysis,” Lecture Notes in Networks and Systems, vol. 750 LNNS, pp. 344–353, 2023, doi: 10.1007/978-3-031-42536-3_33/SAVE-RESEARCH.

[19] O. Mishkov, K. Zorin, D. Kovtoniuk, V. Dereko, and I. Morgun, “Comparative Analysis of Normalizing Techniques Based on the Use of Classification Quality Criteria,” Lecture Notes on Data Engineering and Communications Technologies, vol. 77, pp. 602–612, 2022, doi: 10.1007/978-3-030-82014-5_41.

[20] Y. Sanggelorang, F. Ari Anggraini Sebayang, N. S. H. Malonda, and A. A. Rumayar, “Insights into Childhood Malnutrition: An Analysis on Food Vulnerability and Stunting using 2021 Indonesian Nutritional Status Survey Data,” Media Gizi Indonesia, vol. 19, no. 3, pp. 282–290, Sep. 2024, doi: 10.20473/MGI.V19I3.282-290.

[21] H. Anastasia et al., “Determinants of stunting in children under five years old in South Sulawesi and West Sulawesi Province: 2013 and 2018 Indonesian Basic Health Survey,” PLoS One, vol. 18, no. 5 May, May 2023, doi: 10.1371/JOURNAL.PONE.0281962.

[22] M. R. Ara, “ASSESSMENT OF VULNERABILITY AND SUSTAINABLE LIVELIHOOD OF THE URBAN POOR: A STUDY IN KHULNA CITY CORPORATION, SOUTHWEST BANGLADESH,” Khulna University Studies, pp. 217–228, Oct. 2022, doi: 10.53808/KUS.2017.14.1AND2.1702-S.

[23] D. F. Polit, C. T. Beck, and S. V. Owen, “Focus on research methods: Is the CVI an acceptable indicator of content validity? Appraisal and recommendations,” Res. Nurs. Health, vol. 30, no. 4, pp. 459–467, Aug. 2007, doi: 10.1002/nur.20199.

[24] D. F. Polit and C. T. Beck, “The content validity index: are you sure you know what’s being reported? Critique and recommendations,” Res. Nurs. Health, vol. 29, no. 5, pp. 489–497, Oct. 2006, doi: 10.1002/NUR.20147.

[25] P. J. Rousseeuw, “Silhouettes: A graphical aid to the interpretation and validation of cluster analysis,” J. Comput. Appl. Math., vol. 20, no. C, pp. 53–65, Nov. 1987, doi: 10.1016/0377-0427(87)90125-7.

[26] T. Hastie, R. Tibshirani, and J. Friedman, “The Elements of Statistical Learning,” 2009, doi: 10.1007/978-0-387-84858-7.

[27] F. Huettmann et al., “Use of Machine Learning (ML) for Predicting and Analyzing Ecological and ‘Presence Only’ Data: An Overview of Applications and a Good Outlook,” Machine Learning for Ecology and Sustainable Natural Resource Management, pp. 27–61, 2018, doi: 10.1007/978-3-319-96978-7_2.

[28] G. James, D. Witten, T. Hastie, and R. Tibshirani, “An Introduction to Statistical Learning with Applications in R Second Edition,” 2021.

[29] H. Torlesse, A. A. Cronin, S. K. Sebayang, and R. Nandy, “Determinants of stunting in Indonesian children: evidence from a cross-sectional survey indicate a prominent role for the water, sanitation and hygiene sector in stunting reduction,” BMC Public Health 2016 16:1, vol. 16, no. 1, pp. 669-, Jul. 2016, doi: 10.1186/S12889-016-3339-8.

[30] T. Mulyaningsih, I. Mohanty, V. Widyaningsih, T. A. Gebremedhin, R. Miranti, and V. H. Wiyono, “Beyond personal factors: Multilevel determinants of childhood stunting in Indonesia,” PLoS One, vol. 16, no. 11, p. e0260265, Nov. 2021, doi: 10.1371/JOURNAL.PONE.0260265.

[31] A. D. Laksono, R. D. Wulandari, N. Amaliah, and R. W. Wisnuwardani, “Stunting among children under two years in Indonesia: Does maternal education matter?,” PLoS One, vol. 17, no. 7 July, p. e0271509, Jul. 2022, doi: 10.1371/JOURNAL.PONE.0271509.

[32] C. R. Titaley, I. Ariawan, D. Hapsari, A. Muasyaroh, and M. J. Dibley, “Determinants of the Stunting of Children Under Two Years Old in Indonesia: A Multilevel Analysis of the 2013 Indonesia Basic Health Survey,” Nutrients 2019, Vol. 11, Page 1106, vol. 11, no. 5, p. 1106, May 2019, doi: 10.3390/NU11051106.

[33] H. Anastasia et al., “Determinants of stunting in children under five years old in South Sulawesi and West Sulawesi Province: 2013 and 2018 Indonesian Basic Health Survey,” PLoS One, vol. 18, no. 5 May, p. e0281962, May 2023, doi: 10.1371/JOURNAL.PONE.0281962.

[34] A. Murad, F. Faruque, A. Naji, A. Tiwari, M. Helmi, and A. Dahlan, “Modelling geographical heterogeneity of diabetes prevalence and socio economic and built environment determinants in Saudi City - Jeddah,” Geospat. Health, vol. 17, no. 1, Jun. 2022, doi: 10.4081/GH.2022.1055.