<![CDATA[Indonesia faces challenges of uneven population distribution, including in Garut Regency, which has a population of 2,585,607 people. This disparity leads to overcrowding issues in several areas. This study aims to cluster population density based on districts using the K-Means algorithm and the CRISP-DM (Cross-Industry Standard Process for Data Mining) approach. CRISP-DM consists of six main stages: business understanding, data understanding, data preparation, modeling, evaluation, and deployment, which are systematically applied in this research. The data used was obtained from the Central Bureau of Statistics of Garut Regency in 2023, covering population numbers and the area of each district. At the modeling stage, the K-Means algorithm is applied to group districts based on population density similarity. The optimization of the number of clusters was carried out using the elbow method, resulting in the optimal number of three clusters (k=3). Evaluation using the Davies-Bouldin Index (DBI) yielded a value of 0.5794, indicating that the clusters formed have good separation. The clustering results show that cluster 0 includes districts with high density, cluster 1 with medium density, and cluster 2 with low density. The results of this study have the potential to be implemented in regional development planning, assisting the Central Bureau of Statistics (BPS) of Garut Regency and related agencies in formulating policies for equitable development, public service distribution, and more effective infrastructure planning. With population density mapping based on data mining, policies can be more evidence-based, enabling better decision-making in addressing demographic issues in Garut Regency.]]>
Copyrights © 2025
