Performansi Algoritma Clustering K-Means untuk Penentuan Status Malnutrisi pada Balita

Abstract

Malnutrition has always been a global problem and is the main cause of child mortality in developing countries. Nutritional status is a measure of success in meeting the nutritional needs of toddlers which can be seen from the achievement of toddler's weight. The short-term effects of malnutrition on toddlers are physical growth disorders in children, metabolic disorders, brain development disorders, and intelligence disorders. In the long term, malnutrition will cause a decrease in learning outcomes and cognitive abilities, decreased immunity, the risk of developing diabetes, heart and blood vessel disease, obesity, stroke, cancer, and disability in old age. To find out about the growth and development of early childhood, especially in relation to the development of early childhood nutritional values, regular measurements of the height and weight of toddlers from an early age are carried out through Posyandu activities. Toddler height and weight data will be grouped according to the nutritional characteristics of each toddler. In this study, infants will be clustered based on their nutritional characteristics using the K-Means algorithm clustering method. Based on the evaluation results, the number of clusters 3 is more optimal than the number of clusters 5 in the toddler dataset used in the study. It is hoped that the research results obtained can serve as a basis or recommendation material for posyandu throughout Indonesia to determine the nutritional status of the under-five age group, so that the handling is more appropriate and can help prevent malnutrition in under-fives.