Level alert classification of dengue hemorrhagic fever cases in DKI Jakarta with the implementation of the random forest algorithm

Abstract

In 2007, the capital district, DKI Jakarta had one of the worst floods that submerged nearly 60% of the area. One of the causes was heavy rainfall. Two months after the incident, the Governor of DKI Jakarta stated that the capital district was affected by an outbreak of dengue fever. From these two incidents, there are some indications of dengue hemorrhagic fever cases related to weather. In this research, a random forest classification algorithm was used to build a model that could classify the level alert of dengue fever for each district in DKI Jakarta into three categories: safe, moderately safe, and unsafe. The results and the accuracy of the classification model for the level alert of cases of dengue hemorrhagic fever for East Jakarta, North Jakarta, South Jakarta, Central Jakarta, and West Jakarta are 94.51%, 86.81%, 81.32%, 80.22%, and 80.22% respectively.