JTUS, Vol. 02, No. 2 February 2024 60

JTUS, Vol. 02, No. 2 February 2024

E-ISSN: 2984-7435, P-ISSN: 2984-7427

K-Means Algorithm for District/City Classification in West Java

Based on Stunting Toddler Data

to chronic malnutrition, causing them to have a shorter height than they should for their age

(Ruaida, 2018). This nutritional deficiency can begin to occur since the fetus is in the womb and

continues in the early days after the baby is born (Aprilia, 2020). However, stunting is only seen

after the child reaches the age of two. Stunted toddlers have a body length (PB) or height (TB)

that is lower than the standard set by WHO-MGRS (Multicentre Growth Reference Study) 2006 for

their age, or according to the definition of the Ministry of Health (MoH), stunting occurs when the

z-score of children under five is less than -2SD (stunted) and even less than -3SD (severely stunted)

(Kemiskinan, 2017) (Grantz et al., 2018). The age of 24-59 months is considered a critical period in

the effort to ensure quality human resources, especially since the first two years of a child's life

are a golden period for optimal brain growth and development (Nurhayati, 2020). Therefore,

poverty and development. According to a report from the United Nations Children's Emergency

Fund (UNICEF), more than half of stunted children, about 56%, live in Asia, while more than a third,

to child nutrition and growth problems (Hanifa & Mon, 2021)(Izani, 2021). UNICEF data shows

that around 80% of stunted children are spread across 24 developing countries in Asia and Africa.

Indonesia ranks fifth in the list of countries with the highest stunting prevalence, after India, China,

Nigeria, and Pakistan. Currently, the prevalence of stunting in children under the age of 5 in the

South Asia region reaches around 38%.

Nutrition issues remain a major focus in Indonesia, especially in the context of nutrition in

toddlers. The health condition and nutritional status of toddlers are important indicators of overall

public health. This is due to the impact caused by cases of malnutrition, undernutrition, stunting

(growth delay), and other nutritional problems that are a burden for families, communities, and

countries (Unicef, 2012). Some factors that are suspected of causing stunting include the mother's

development that is not optimal in children, increased morbidity and mortality, body posture that

is not optimal in adulthood (shorter than average), and learning capacity and performance that is

less than optimal at school (Organization, 2020). Cluster analysis is one of the important methods

in the field of Data Mining. Data mining is a process that uses various statistical, mathematical,

artificial intelligence, and machine-learning techniques to extract and identify valuable

information from various large databases (Sitepu et al., 2011). Cluster analysis in data mining is a

method used to group a series of data into groups based on predetermined similarities (Matdoan

& Van Delsen, 2020). Among the various cluster analysis methods available, two of them are K-

Means and K-Medoids Clustering. Both of these methods are types of partitioning clustering

methods that have interrelated algorithms. Such methods tend to be faster than hierarchical

METHODS

The dataset was obtained from data on the number of stunting toddlers by district/city in

the Health Office which is issued every 1 year.

Figure 2. The process of clustering using the K-Means

algorithm in RapidMiner Studio.

Based on the results of Data Mining testing using the K-Means Clustering algorithm in the

RapidMiner 10 software application, the conclusion that can be drawn is that the centroid value

obtained is different due to differences in the amount of data, but the value in the resulting cluster

remains the same (Sangga, 2018). Details of the final centroid values can be seen in Figure 2 below:

Based on Figure 3, it can be seen that the classification of districts/cities based on the

number of stunting in West Java Province is divided into two clusters. Cluster 0 consists of 17

in Table 2 below.

Table 2. Cluster List

Cluster 1

10

Kabupaten Bogor, Kabupaten Sukabumi, Kabupaten Cianjur,

Kabupaten Bandung, Kabupaten Garut, Kabupaten Tasikmalaya,

Kabupaten Cirebon, Kabupaten Indramayu, Kabupaten Karawang, Kota

Bekasi

The second group (cluster 1) consists of Bogor Regency, Sukabumi Regency, Cianjur Regency,

Bandung Regency, Garut Regency, Tasikmalaya Regency, Cirebon Regency, Indramayu Regency,

Karawang Regency, Bekasi City.

Ahmed, M., Seraj, R., & Islam, S. M. S. (2020). The k-means algorithm: A comprehensive survey and

performance evaluation. Electronics, 9(8), 1295.

Aprilia, W. (2020). Perkembangan pada masa pranatal dan kelahiran. Yaa Bunayya: Jurnal

Pendidikan Anak Usia Dini, 4(1), 39–56.

Grantz, K. L., Hediger, M. L., Liu, D., & Louis, G. M. B. (2018). Fetal growth standards: the NICHD

fetal growth study approach in context with INTERGROWTH-21st and the World Health

Overcoming Child Trafficking In China. Universitas Muhammadiyah Yogyakarta.

Kemiskinan, T. N. P. P. (2017). Pemilihan 10 Desa Prioritas di 100 Kabupaten/Kota Prioritas

Penanganan Kemiskinan dan Stunting. Jakarta: TNPPK.

Matdoan, M. Y., & Van Delsen, M. S. N. (2020). Penerapan Analisis Cluster Dengan Metode Hierarki

Untuk Klasifikasi Kabupaten/Kota Di Provinsi Maluku Berdasarkan Indikator Indeks

Pembangunan Manusia. Statmat J. Stat. Dan Mat, 2(2), 20.

Nurhayati, R. (2020). Pendidikan Anak Usia Dini Menurut Undang–Undang No, 20 Tahun 2003 Dan

Sistem Pendidikan Islam. Al-Afkar, Journal For Islamic Studies, 57–87.

Organization, W. H. (2020). World health statistics 2020.

Pangestuti, M., Khomsan, A., & Ekayanti, I. (2023). Determinants of stunting in children aged 6-24

months in rural areas: Case control study. AcTion: Aceh Nutrition Journal, 8(3), 318–330.