<![CDATA[Poverty alleviation is one of goals at Sustainable Development Goals (SDGs) that can be supported by grouping poor households based on similar characteristics, enabling social assistance programs to be delivered more effectively and accurately. This study aims to cluster poor households in Samarinda City using the K-Modes algorithm with validation through the Davies-Bouldin Index (DBI). The data were obtained from the 2023 Samarinda Poverty Survey and consist of 16 poor households indicators. The analysis was conducted by testing the number of clusters (K) from 2 to 10. The results indicate that the optimal number of clusters is 3, with a DBI value of 1.6997. Cluster 1 consists of 11,237 households, Cluster 2 consists of 2,327 households, and Cluster 3 consists of 1,451 households. The distinct characteristics of each cluster suggest that the clustering results can serve as a basis for designing more targeted social assistance programs. Future research is recommended to consider alternative clustering methods for categorical data, such as the ROCK algorithm, which utilizes link-based similarity by considering the number of common neighbors between objects, allowing it to better capture the inherent structure of categorical data compared to distance-based methods.]]>
Copyrights © 2026
