Implementasi Algoritma Naive Bayes Dengan Feature Selection Backward Elimination Dalam Pengklasifikasian Status Penderita Stunting Pada Balita

Abstract

Stunting or stunting is one of the nutritional problems experienced by toddlers, where toddlers experience failure to thrive as a result of chronic malnutrition so that toddlers are too short for their age. Broadly speaking, stunting is caused by a lack of nutrition for a long time and the occurrence of recurrent infections, and these two causative factors are influenced by inadequate parenting from the womb to the first 1,000 days of birth. The Asian Development Bank (ADB) reports that the prevalence of children with stunting under the age of five in Indonesia is the second highest in Southeast Asia. Its prevalence reaches 31.8% in 2020. Further monitoring and data collection by the Singgahan Pukesmas regarding stunting cases determines the growth and development factors of toddlers both in the womb and toddlers who have been born. However, the problem that often arises at the Singgahan Pukesmas is that examining the status of stunting in toddlers still takes quite a long time because it is done manually and is also prone to inaccuracies, so a system is needed that can classify toddler examination data to predict whether the child is in stunting or not stunting status. fast and accurate. From the results of this study it can be concluded that the Naive Bayes Algorithm with backward elimination feature selection makes it easier to determine the status of stunted or not stunted toddlers with the variables gender, age, weight, height, BB/U, Z-core BB/U, BB/ TB, Z-Core BB/TB, Z-core TB/U with a total of 450 dataset records, 360 training data records and 90 testing data records taken randomly with an accuracy of 86.11%