Comparative Analysis of Machine Learning Algorithms for classification about Stunting Genesis

Abstract

. Background The use of machine learning is very much needed for health experts as data and information processing to make it easier to analyze automatically. To produce accuracy in solving problems. Application of machine learning with comparative three algorithms to solve stunting problems. Because toddlers in Indonesia are still high, especially at age 2 -3 years. Seen from many factors that are at risk of causing stunting. The instrument is needed in Machine Learning. The goal (1). In addition to providing knowledge in the field of Informatics. It’s also useful for health experts in managing data in making decisions, as to facilitate analysis automatically. (2). Can reduce the impact on the incidence of stunting. Methods Comparison of three algorithms in the classification of the results. That was compared yielded an accuracy of 86% AUC 0.85 for the Decision Tree algorithm with a diagnosis level of Good classification, Algorithm KNN with an accuracy of 58.7% AUC 0.57 fail classification, Algorithm Naïve Bayes with 55% AUC accuracy 0.51, using 13 stunting data variables.