Analysis Of Poverty Level Mapping In Riau Province Using The K-Means Method

Abstract

This research aims to understand the pattern of poverty distribution in Riau Province, identify clusters that reflect similar characteristics and provide a basis for developing more targeted policies. This approach uses machine learning techniques, especially the K-Means algorithm, to form clusters based on poverty level data. The results of the analysis show cluster 0 (C0) with a high poverty level and cluster 1 (C1) with a low poverty level. K-Means proved effective in grouping areas with similar levels of poverty, and provided a strong foundation for further analysis. Evaluation results using the Adjusted Rand Index (ARI), Silhouette Score, Davies-Bouldin Index, and Calinski-Harabasz Index show that the quality of cluster formation is good. This analysis provides detailed insight into poverty patterns in Riau Province and provides an empirical basis for implementing more contextual policies.