<![CDATA[This study aims to classify underprivileged communities in Kelurahan Durian, Medan, to optimize social assistance distribution using the OPTICS algorithm. The socio-economic data used includes income, expenditure, occupation, education level, and number of dependents, comprising 800 records, which after preprocessing became 789 data points. The research stages include preprocessing, parameter determination through K-Distance Plot and Grid Search, the clustering process, and evaluation using the Silhouette Index. Optimal parameters were obtained at min_samples = 15, max_eps = 0.3, and xi = 0.030, yielding a Silhouette Index value of 0.2409. The clustering produced 4 clusters: unable, underprivileged, capable, and highly capable, along with a number of noise points. The OPTICS algorithm proved effective in identifying data structures with varying densities and automatically detecting outliers. Results were visualized through a reachability plot. This study is expected to improve the accuracy of targeted social assistance distribution through a data-driven approach. Keywords: Clustering, OPTICS, Data Mining, Social Assistance, Poverty]]>
Copyrights © 2026
