The decision tree classification with C4.5 and C5.0 algorithm based on R to detect case fatality rate of dengue hemorrhagic fever in Indonesia

Abstract

Abstract Dengue hemorrhagic fever (DHF) is a deadly disease that is transmitted through mosquito bites from the genus Aedes especially Aedes aegypti, Aedes aegypti can occur every year and affect any age. DHF has a high case fatality rate (CFR). Therefore we need a method that can detect CFR of DHF in Indonesia, one of which is the decision tree classification based on C4.5 and C5.0 algorithm. C4.5 and C5.0 algorithm starting with forming a root node and ending with a leaf node by evaluating attributes using information gain to measure the effect of attributes in classifying a dataset. In this article, an applied research is carried out, namely applying the decision tree classification with C5.0 and C4.5 algorithm based on R software to detect CFR of DHF in Indonesia. The attributes used are the incidence rate (IR), population density, many hospitals, and many medical personnel. The results show that C5.0 algorithm has a big error than the C4.5 algorithm, while the C5.0 algorithm has a smaller tree than the C4.5 algorithm.