CD Skripsi
Implementasi Clustering Jumlah Penduduk Berdasarkan Umur Menggunakan Metode K-Medoids (Studi Kasus : Disdukcapil Provinsi Riau)
The rapid growth of the population can pose various challenges in managing civil administration, including public policy planning. Therefore, an effective data analysis method is needed to cluster the population based on age. This study aims to implement the K-Medoids method in the clustering process of the population based on age at the Office of Population and Civil Registration of Riau Province. The K-Medoids method was chosen for its ability to produce more stable centroids compared to the K-Means method, especially when dealing with uneven data. The data used consists of the total population of Riau Province for the years 2023 and 2024. The analysis process involves several stages, including data collection, preprocessing, and data normalization. The analyzed indicator variables include age, male, female, and total population. The study results indicate that the population can be grouped into two optimal clusters based on the Elbow Method. In 2023, the obtained Silhouette Index value was 0.7708, which increased to 0.7789 in 2024, indicating a very strong clustering quality (Strong Structure). Further analysis shows that the population aged 1–59 years increased by 3.5%, while the population aged 60–121 years experienced a 6.4% increase. These findings demonstrate that the K-Medoids algorithm can effectively cluster the population based on age and assist the government in formulating public service policies that align with the needs of each age group.
Keywords: Clustering, Data Mining, Population, K-Medoids, Silhouette Index
Tidak tersedia versi lain