COMPARISON OF FUZZY C-MEANS AND FUZZY GUSTAFSON-KESSEL CLUSTERING METHODS IN PROVINCIAL GROUPING IN INDONESIA BASED ON CRIMINALITY-RELATED FACTORS

Indonesia is a country that has population density increasing every year, with the increase in density, crime rate increasing. Criminal acts arise because they are supported by factors cause crime. To improve security and welfare of Indonesian people, authors grouped each province based on influence This study uses comparison Fuzzy C-Means Clustering (FCM) Gustafson-Kessel (FGK) methods using validation index for determining optimal cluster, namely Davies Bouldin Index The data used secondary form variables forming affect Indonesia, where obtained comes from website Central Statistics Agency (BPS). results this FGK method better than FCM have smaller standard deviation ratio. grouping best method, FGK, it was found number clusters formed 5 cluster 1 consisting 6 provinces, 2 4 3 11 8 provinces.