Cluster Analysis of Indonesian Provinces Based On Harvest Area And Rice Productivity Using Single Linkage Method

Abstract

In this article, a cluster analysis will be conducted for provinces in Indonesia based on the harvest area (ha) and rice productivity (ku/ha) of 34 provinces in Indonesia. Clustering is done using a hierarchical method, namely single linkage. The distance used as the basis for clustering is the euclidian distance. Based on the results of clustering using single linkage, 3 large clusters were obtained. In this article, a cluster analysis will be conducted for provinces in Indonesia based on the harvest area (ha) and rice productivity (ku/ha) of 34 provinces in Indonesia. clustering is done using a hierarchical method, namely single linkage. The distance used as the basis for clustering is the euclidian distance. Based on the results of clustering using single linkage, 3 large clusters were obtained. Cluster consists of 3 provinces, cluster 2 consists of 1 province and cluster 3 consists of 30 provinces. Cluster 1 is a province with high rice production with an average total rice production of 9,628,788 tons. Cluster 2 with an average total rice production of 5,341,021 tons. While cluster 3 with an average rice production of 863,995.34 tons. Furthermore, based on cluster validation using the anova test, the significance value is 0.00>0.05, which means that there is a significant difference between clusters. Thus it can be stated that the division of 34 Indonesian provinces in terms of land area and rice productivity into 3 large clusters using the single linkage method is valid.