Implementasi Metode Bagging dan Teknik Discretization pada Algoritma Machine Learning untuk Memprediksi Status Stunting pada Anak Balita

Abstract

Stunting is one of the problems of toddler growth, making toddlers susceptible to disease. Efforts to prevent stunting with routine checks every month. Posyandu Sukasejati is a facility for routinely checking the growth of toddlers, but data collection requires early stunting analysis to help health workers reduce the number of stunting statuses in toddlers. Previous research on stunting prediction using the Naive Bayes Machine Learning algorithm has been carried out, but the level of accuracy is still low, so accuracy improvement techniques are needed to provide accurate information. The purpose of the study was to implement the Bagging method to improve accuracy and the discretization technique to change continuous attributes to categorical in the Naïve Bayes Machine Learning algorithm in predicting stunting in toddlers, the results of the study showed an increase in accuracy, recall, and precision using a combination of the Bagging method and the Naïve Bayes algorithm, namely accuracy of 100% increased by 5.83% compared to using the Naïve Bayes algorithm alone, which was 94.17% and an increase in recall and precision results of 28.33%.