Optimization of Davies-Bouldin Index with k-medoids algorithm

This study will employ the Davies Bouldin Index (DBI) technique to determine the number of clusters to be formed, with the goal of maximizing the number of clusters formed. The method used in this study is k-medoids, which are evaluated using the DBI technique prior to application. The data for this...

Full description

Saved in:
Bibliographic Details
Main Authors: Henderi, H., Fitriana, Liza, Iskandar, I., Astuti, Rina, Arifandy, M. Imam, Hayadi, B. Herawan, Mesran, M., Chin, Jacky, Kurniawan, Andi
Format: Conference Proceeding
Language:English
Subjects:
Algorithms
Clusters
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This study will employ the Davies Bouldin Index (DBI) technique to determine the number of clusters to be formed, with the goal of maximizing the number of clusters formed. The method used in this study is k-medoids, which are evaluated using the DBI technique prior to application. The data for this study were obtained from the Indonesian Central Statistics Agency’s (BPS) official website, which can be accessed at the following URL: https://bps.go.id. The data for this study comes from Indonesia’s rice production in 2019-2020, which spans 34 provinces. The data is processed using RapidMiner software version 5.3’s Performance feature, specifically the Davies-Bouldin Index technique. To obtain the best cluster results, begin testing from cluster = 2 to cluster = 10. Cluster = 4 has the highest number of cluster formations, with a DBI value of -1.413. Thus, four clusters were formed, with cluster 1 comprising 22 provinces, cluster 2 comprising three provinces, cluster 3 comprising two provinces, and cluster 4 comprising seven provinces.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0225220