Toddler nutritional status identification: Support vector machine (SVM) algorithm adoption

Affan Asyraffi; Okfalisa Okfalisa; Fitri Insani; Surya Agustian; Riski Mai Candra

doi:10.59190/stc.v5i2.282

Authors

Affan Asyraffi Department of Informatics Engineering, UIN Sultan Syarif Kasim Riau, Pekanbaru 28293, Indonesia
Okfalisa Okfalisa Department of Informatics Engineering, UIN Sultan Syarif Kasim Riau, Pekanbaru 28293, Indonesia
Fitri Insani Department of Informatics Engineering, UIN Sultan Syarif Kasim Riau, Pekanbaru 28293, Indonesia
Surya Agustian Department of Informatics Engineering, UIN Sultan Syarif Kasim Riau, Pekanbaru 28293, Indonesia
Riski Mai Candra Department of Informatics Engineering, UIN Sultan Syarif Kasim Riau, Pekanbaru 28293, Indonesia

DOI:

https://doi.org/10.59190/stc.v5i2.282

Keywords:

Classifications, Data Mining, Nutritional Status, Support Vector Machine, Toddlers

Abstract

Inadequate nutrition in toddlers can lead to health issues and adversely affect their growth, development, and cognitive capabilities. Consequently, it is essential to assess the nutritional status of toddlers to ascertain their health level. This study seeks to ascertain the nutritional health of toddlers utilizing the support vector machine (SVM) methodology, taking into account body weight (BB), height (TB), age, BB/TB ratio, Z-scores for BB/U, Z-scores for TB/U, and Z-scores for BB/TB. The data of 1458 toddlers were evaluated using the knowledge data discovery methodology. This study effectively categorized toddler nutrition into six classifications including malnutrition, undernutrition, adequate nutrition, overnutrition, risk of overnutrition, and obesity. Utilizing the confusion matrix methodology with an 80% training data to 20% test data ratio yields an accuracy of 89.04%. The SVM method is effectively utilized to ascertain the nutritional condition of toddlers, hence enhancing their growth and development.