Classification of Stunting Events: Case Study in West Java, Indonesia

This study created a categorization model based on stunting episodes, separating stunting and non-stunting groups. Stunting is a persistent nutritional condition in toddlers who are shorter in height than other children their age. Data mining is one of the strategies used to extract knowledge from a large amount of big data and convert it into fresh data that can be understood as technology advances. The categorization model results provide a foundation for mapping the possibility of stunting events using the following methods: K-Nearest Neighbour, Logistic Regression, and Support Vector Machine (SVM). The proposed categorization model identifies stunting in toddlers from Pandeglang Regency, West Java, Indonesia, and Malaysia. The sample taken in this study consists of 798 children data. This study concludes that the three tests were successfully carried out, with data sharing ratios of 80%: 20%, 75%: 25%, and 70%: 30%, respectively. The model was repeated 100 times using a Support Vector Machine (SVM) and Logistic Regression (LR) and evaluated for precision, recall, f-1 score, and accuracy. The results of three experiments revealed that the best method is LR utilizing the chi-squared test feature selection, which achieved an accuracy of 85%.