K-Means Clustering Algorithm for Determination of Clustering of Bangkalan Regional Development Potential. (July 2020)

• Record Type: Journal Article
• Authors: Doni, A F; Negera, Y D P; Maria, O A H
• Is Part Of: Journal of physics. Volume 1569:Number 2(2020)
• Publication Date: 2020-07
• Subjects: Clustering -- K-Means -- Regional Potential -- Preprocessing -- Sector -- Scaling; Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1569/2/022078
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: Regional potential is the ability of an area that can be developed to improve the area according to its regional characteristics. This potential can cover social, economic, infrastructure, natural resources, and human resources and their work. To improve regional development, a strategic policy is needed that leads to improving the economy of a region. In this study the regional grouping of 3 agricultural, plantation, and livestock groups in Bangkalan Regency consisted of 18 sub-districts using the K-Means Clustering method, from which the data was processed preprocessing using Scaling transformations starting from the lower limit search and the upper limit, looking for the Lower Limit and Upper Limit, then grouping using K-Means Clustering. From the research, it was found that testing in the agricultural sector with 3 clusters for the initial cluster center 5, 7, 15 obtained final clusters according to the highest yield criteria in cluster 1 compared to clusters 2 and 3. For testing in the plantation sector with 3 clusters for the initial cluster center 4, 7, 16 the final clusters were obtained according to the highest yield criteria in cluster 2 compared to clusters 1, and 3 and for testing the livestock sector with 3 clusters at the initial cluster center 3, 11, 13 the final clusters were obtained according to the criteria it was known that more height in cluster 2 is compared to cluster 1 and 3.