APLIKASI NAÏVE BAYES CLASSIFIER (NBC) PADA KLASIFIKASI STATUS GIZI BALITA STUNTING DENGAN PENGUJIAN K-FOLD CROSS VALIDATION

Abstract

The case of stunting in Indonesia is a problem that has been discussed for a long time. One of many efforts to overcome this problem is through an accelerated stunting reduction program to improve the nutritional status of the community and also to reduce the prevalence of stunting or stunted toddlers. Generally, the index used to determine the nutritional status of stunting toddlers height compared to age. This study aims to identify the classification results, evaluate the model, and predict the nutritional status of stunting toddlers using the Naïve Bayes Classifier algorithm with K-Fold Cross Validation testing. The data processing system used is the GUI-R (Graphical User Interface) in order to facilitate the analysis process by implementing the Shiny Package in the Rstudio program. The results of accuracy using Naïve Bayes Classifier with 10-Fold Cross Validation test obtained the highest accuracy on the 6th iteration with an accuracy 94.39%, while the lowest accuracy on the 8th iteration with an accuracy 82.08%. Overall, the average accuracy in each iteration is 88.46%, so it can be concluded that Naïve Bayes Classifier model considered good enough to classified data on the nutritional status of stunting toddlers.

Keywords: Stunting, Data Mining, Naïve Bayes Classifier, K-Fold Cross Validation, Shiny Package