Journal of Computer Science, Information Technology and Telecommunication Engineering (JCoSITTE)
Vol 4, No 2 (2023)

Implementation of Data Classification Using K-Means Algorithm in Clustering Stunting Cases

Indah Purnama Sari (Universitas Muhammadiyah Sumatera Utara)
Al-Khowarizmi Al-Khowarizmi (Universitas Muhammadiyah Sumatera Utara)
Oris Krianto Sulaiman (Universitas Islam Sumatera Utara)
Dicky Apdilah (Universitas Asahan)



Article Info

Publish Date
31 Aug 2023

Abstract

Stunting is still a serious public health problem in Indonesia, where the prevalence of this condition is 37.2%, up from 35.6% in 2019 and 36.8% in 2020. The length or height of a child who is short (dwarf) is below average for his age. Stunting has a negative impact on IQ deficiencies, infectious diseases, mental health problems, and child development. Toddlers with stunting cases are detected when their growth and development does not match their age, but currently there is no data grouping based on these criteria that is of concern to parents and posyandu cadres. Data can be grouped using the K-Means data mining technique. The K-Means algorithm is often used by researchers as a grouping procedure to ascertain whether children are stunted or not. 395 datasets are used in this research data. The Knowledge Discovery In Databases (KDD) approach, a comprehensive nontrivial procedure for detecting and recognizing patterns in data, underlies this research. Based on the variables of age, weight and height, this study aims to identify groups or clusters of stunting status in children under five. The best number of clusters with K = 2 was determined by the findings of this investigation. There are 392 children in cluster 0-Shanum, Rizka, Nurjanah, and others-and three toddlers in cluster 1-Ezra, M. Abidza, and Abd Mahmud. With a total of 287 stunted toddlers and 108 toddlers with normal development, the most ideal DBI value is 0.007 which is close to 0, this shows that the clusters under review provide quality clusters.

Copyrights © 2023






Journal Info

Abbrev

jcositte

Publisher

Subject

Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering

Description

ournal of Computer Science, Information Technology and Telecommunication Engineering (JCoSITTE) is being published in the months of March and September. It is academic, online, open access (abstract), peer reviewed international journal. The aim of the journal is to: Disseminate original, ...