APLIKASI PREDIKSI STATUS GIZI BALITA BERBASIS WEB MENGGUNAKAN METODE K-NEAREST NEIGHBOR

Abstract

The nutritional status of children under five is an important parameter in monitoring and caring for children's health, and this research aims to provide a reliable and efficient tool for stakeholders, including health workers and parents. The K-NN method was chosen as the basis for prediction because of its ability to classify nutritional status by comparing the similarity of new toddler data with existing data in the dataset. This application allows users to enter data such as the age, weight, height and gender of the toddler. Using the K-NN algorithm, this application automatically calculates the distance between the toddler to be predicted and the toddlers in the dataset that are most similar to it. Application testing results show a very satisfactory accuracy level of 91.94%. This shows that this application can provide predictions of the nutritional status of toddlers with a high level of confidence, potentially enabling early detection of nutritional problems and timely intervention.