Classification of Pneumonia Risk Factor Characteristics in Toddlers Using Classification and Regression Trees (CART) Case Study Regional General Hospital dr. Tengku Mansur Tanjungbalai

Mangunsong, Elvira Yolanda
Huda Prasetya, Nurul
Rakhmawati, Fibri

Quadratic: Journal of Innovation and Technology in Mathematics and Mathematics Education
Vol. 1 No. 2 (2021): October 2021

Pneumonia is one of the leading causes of death in children in the world. The World Health Organization (WHO) estimates that this disease causes 16% of deaths in children under 5 years of age worldwide. Pneumonia is included in the top 10 diseases that suffer the most every month at the RSUD dr.Tengku Mansyur Tanjungbalai, throughout 2019 there were 73 children under five years with pneumonia and 5 of them were declared dead without hospitalization. This study is useful to find out the results of the classification accuracy of the classification and regression trees (CART) for risk factors for pneumonia in children under five years. CART classification is done by dividing the total data of under-five patients with the ratio of 15% testing data and 85% learning data. The classification accuracy obtained for the prediction data was 50% with the percentage of sensitivity and specificity of 94,12% and 14,29% respectively.