Komparasi K-Means Clustering dan K-Medoids Clustering dalam Mengelompokkan Produksi Susu Segar di Indonesia Berdasarkan Nilai DBI

Mochamad Wahyudi
Solikhun Solikhun
Lise Pujiastuti

Jurnal Bumigora Information Technology (BITe)
Vol 4 No 2 (2022)

The purpose of this study was to find the optimal grouping from the comparison of the two methods in grouping fresh milk production using the K-Means algorithm and the K-Medoids algorithm. To find optimal grouping, the authors compare the grouping results by looking for the smallest DBI (Davies Bouldin Index) value. The data used in this study is data on fresh milk production in Indonesia which is sourced from the Indonesian Central Bureau of Statistics for 2018-2020. Evaluation of the DBI value for the K-Means Clustering algorithm is 0.094 and the DBI value for K-Medoids Clustering is 0.072. Therefore, grouping fresh milk production using the K-Medoids algorithm has better results than using the K-Means Clustering algorithm, because the K-Medoids Clustering algorithm has a smaller DBI value of 0.072. The benefit of this study is to obtain optimal clusters in classifying fresh milk in Indonesia to provide information to the government in increasing fresh production in Indonesia in the future.