APPLICATION OF K-MEANS METHOD FOR CLUSTERING COFFEE EXPORT DATA

Abstract

Data on coffee exports from 2000 to 2020 by the Directorate General of Customs and Excise through the website of the Central Statistics Agency based on net weight (net) and Free On Board (FOB) values are grouped into 3 clusters with the aim of knowing the amount of coffee production exported by destination country, highest priority for coffee export activities and knowing the marketing potential for coffee to destination countries. The K-Means Clustering method was used to cluster coffee export data, after four iterations the results were obtained, namely the United States was included in the high export volume cluster with a centroid of 1,273,017, Japan, Germany, and Italy including a medium export volume cluster with a centroid of 871,607, and Singapore, Malaysia, India, Egypt, Morocco, Algeria, England, Romania, Georgia, Belgium, the Netherlands, Denmark, and France are included in the low export volume cluster with a centroid of 204,979.