ANALISIS KLASTER PELANGGAN LISTRIK BERDASARKAN PERILAKU KONSUMSI DI KOTA SUKABUMI MENGGUNAKAN METODE K-MEANS CLUSTERING

Rapid population and economic growth in urban areas demands a deeper understanding of customer electricity consumption patterns. This study aims to classify electricity customers in Sukabumi City based on their consumption behavior using the K-Means clustering method. The analysis process was carried out on a dataset that includes Installed Power (DAYA), Energy Consumption (PEMKWH), and On-Time Hours (JAMNYALA). To determine the optimal number of clusters, the Elbow and Silhouette methods were used. The analysis results successfully identified three distinct and consistent customer clusters: the Low Cluster dominated by household customers with the most economical usage, the Medium Cluster indicating normal and stable consumption patterns of household customers, and the High Cluster representing business or industrial segments with very intensive energy consumption. Although the 'High' cluster was successfully identified, its number was not dominant, indicating that the majority of customer data in this period was dominated by segments with low to medium consumption. The conclusion of this study is that K-Means clustering is effective in revealing hidden customer segments. These insights are invaluable for formulating business strategies, such as more personalized product offerings, targeted energy efficiency programs, and more accurate electricity load management planning.
Keywords: Clustering, K-Means, Consumption Behavior, Electricity Customers, Data Analysis.