The Spreading Prediction of Dengue Hemorrhagic Fever (DHF) in Bandung Regency Using K-Means Clustering and Support Vector Machine Algorithm

Abstract

Dengue Hemorrhagic Fever (DHF) is the health problem that exist in tropical country, includes Indonesia. Especially for the Bandung Regency, DHF sufferers fluctuated in the last three years. Through data by Health Department of Bandung Regency recorded from 2014 to 2016, in 2014 recorded as many as 524 cases, in 2015 as many as 1,017 cases, and then in 2016 as many as 3476 cases. Many factors that cause people become DHF sufferers in Bandung Regency are constantly increasing, some of them are high rainfall and also lack of awareness of the cleanness. In this research presents the research about the prediction of DHF in Bandung Regency using K-Means Clustering as preprocessing method and Support Vector Machine (SVM) algorithm as classification method according to historical data of DHF and weather data from BMKG (Meteorological, Climatological, and Geophysical Agency) in Bandung Regency from 2009 until 2016 using the dot and radial kernels on the SVM algorithm. The radial kernel obtains testing accuracy up to 93%, while the kernel dot obtains average of testing accuracy 62%.