Implementation of clustering unsupervised learning using K-Means mapping techniques

Abstract

Abstract The manufacturing sector is one of the major contributors to the Indonesian economy. Human work is still needed on the production floor in the manufacturing industry to ensure a smooth operation. This study explores the use of unregulated learning clustering techniques in data mining in the form of clusters of employees in the production industry. The data collection process is carried out through a survey conducted by the Central Statistical Agency (abbreviated as the BPS) with Url: https:/www.bps.go.id in Large Medium Industries and Micro & Small Industries. The statistics used include 24 industrial classifications, with the number of manufacturing employees in the 2017-2019 industry as a percentage. The unregulated technique of learning clustering is k-means. The Large Cluster (E1) and the Low Cluster are the two labels used (E2). The Davies Bouldin Index (DBI) parameter with a dbi value of 0,929 was used to evaluate the cluster (k=2). The findings showed five manufacturing sectors of the high cluster in 5 cluster and 19 manufacturing sectors of the small cluster in 0 cluster. For each cluster the centroid value is 1.67; 1.64; 1.592 (cluster 1/E1) and 0.348; 0.343; 0.3447 (cluster 0/E2), respectively. The research findings will inform the government to improve labour absorption, which will reduce the unemployment rate by substantial numbers in each manufacturing industry.