The purpose of this research is to utilize data mining techniques in classifying the cost of tertiary education in Indonesia by region. The data source was obtained from the National Socio-Economic Survey (Susenas) Module of Socio-Culture and Education which was managed by the Central Statistics Agency (abbreviated as BPS) in the 2017/2018 school year consisting of 35 regions in Indonesia. The variable used is the average tuition. The method used is K-Medoids which is part of clustering. The data processing is assisted with the RapidMiner 5.3 application. Cluster label determination using Cluster Distance Performance tools with Davies Bouldin performance = 0.428. So that the cluster label used 3 is C1: high cluster, C2: normal cluster and C3: medium cluster. The results of the study stated that 6 regions were in the high cluster (C1) for the cost of education at the tertiary level (Bali, Banten, DI Yogyakarta, DKI Jakarta, West Java, South Sulawesi); 18 areas are in the normal cluster (C2) and 11 areas are in the lower cluster (C3). It is expected that regions that have higher education costs at the tertiary level can be of particular concern to the government, because to improve the quality of human resources in Indonesia, it still requires a significant amount of funding.
Volume 12 | Issue 6
Pages: 2084-2089
DOI: 10.5373/JARDCS/V12I6/S20201169