<![CDATA[Diarrhea is one of the leading causes of morbidity among toddlers in Indonesia. Environmental factors such as drinking water quality, sanitation, maternal hand hygiene, and immunization status contribute significantly to the incidence of diarrhea. This study aims to analyze the application of the Random Forest algorithm in developing a predictive model for diarrhea in toddlers using secondary data from a community health center (Puskesmas), consisting of 200 records divided into 150 training data and 50 testing data. The model was constructed by generating multiple decision trees and combining them using a majority voting technique. The results show that the Random Forest algorithm achieved an accuracy of 88%, precision of 77.78%, recall of 87.5%, F1-score of 82.35%, and specificity of 88.24%. These values indicate that Random Forest is quite reliable in detecting positive diarrhea cases, although some limitations remain in reducing misclassification of negative data. This study contributes to the utilization of machine learning algorithms, particularly Random Forest, as a decision-support tool in the health sector for diarrhea prevention among toddlers.]]>
Copyrights © 2025
