Evaluating the K-Means Analysis in Clustering Area Based on Estates Productivity in Tana Luwu Using Silhouette Index

Abstract

Abstract Tana Luwu is an area in South Sulawesi Province that consists of 4 districts or city, namely Luwu, East Luwu, North Luwu, and Palopo. Those regions have abundant natural resources, one of which is plantation crops. In this manuscript, the first objective was to group the 45 sub-districts in those districts into 2 to 5 clusters based on their productivity on providing plantation crops using K-Means Algorithm. Variables used in this study are coconut, palm, coffee, pepper, cocoa, cloves, and land area. Secondly, after the clusters were formed, those were evaluated using Silhouette Index to find out the most optimum number of clusters. As the result, it was found that the optimum number was two with Silhouette Index 0,8068 while the value of 3 clusters is 0,6468. For the values of 4 clusters and 5 clusters are 0,5029 and 0,5509 respectively.