K-MEANS CLUSTERING UNTUK PEMETAAN DAERAH RAWAN DEMAM BERDARAH

Abstract

District Nogosari is one of the dengue-prone areas in Boyolali District. During the period of 2012 to 2014, there was a significant increase in dengue cases at Boyolali district. For the reasons above, the study is focused on how to cluster Areas DHF-Prone using K-Means method. Clustering is based on the parameter of the number of dengue cases in the sporadic and endemic zones. There are several types of data collection methods that include: observation, interview, and literature study. Design of this proposed system use modeling language the context diagram and data flow diagram. The system is implemented using PHP Programming Language and MYSQL database. This system cluster 3 level zones of Endemic and 3 level zones of Sporadic based on geographic information systems. The result of system testing using the silhouette coefficient on the sporadic zone is the average coefficient for level 1 is 0.837, level 2 is 0.858, and level 3 is 0.773 that means the object has been in the right group. The proposed system is expected to be a consideration in preventing, controlling and eradicating dengue hemorrhagic fever.

Keywords: Endemic, Sporadic, K-Means clustering, Dengue hemorrhagic fever.