Pengelompokkan Kabupaten/Kota di Provinsi Nusa Tenggara Timur Berdasarkan Indikator Indeks Pembangunan Manusia Menggunakan K-Medoids Clustering

Abstrak

The Human Development Index (HDI) is an indicator to measure success in improving the quality of human life. The utilization of HDI indicators can be used to classify observations into clusters based on several aspects, such as health, education, and economy. The results of the clusters can be used as a reference for evaluating government policies. This study uses HDI indicator data in the East Nusa Tenggara (NTT) Province, which consists: Life Expectancy, Average Years of Schooling, Expected Years of Schooling, and Per Capita Expenditure. The outliers were detected in each variable. Therefore the k-medoids clustering method was used in this study because of their robustness on outliers. The clustering results show that there are 4 clusters formed, with each cluster describing a unique character. Cluster 1 describes the condition of the districts in general. Cluster 2 describes the condition of the districts that are one step ahead of others, which are expected to transform into cities in the future. Cluster 3 describes the condition of the district that needs more attention from the government in all sectors because all the HDI indicators are on the lowest value comparing to the other clusters. On the other hand, cluster 4 describes urban areas which have more developed than other areas, in terms of health, education, and economy. The clustering results can be used as a suggestion for evaluating government policies that have been implemented, in order to create structured equitable development based on data and regional characteristics as well.