Grouping of Electricity Regions with K-Medoid Algorithm

Abstract

All human activities are inseparable from the current use of electricity. Electricity sourced from renewable energy is an alternative for areas that have difficulty accessing electricity from PLN. This research gathers data on electricity consumers in Indonesia by utilizing the K-Medoid algorithm and the Davies-Bouldin Index (DBI). The DBI is used to test the quality of clustering so that the optimal cluster in this study is obtained with the lowest DBI value of 0.386 with 3 clusters. The cluster analysis results reveal that fewer areas utilize electricity sourced from PLN compared to non-PLN or renewable energy sources. Based on this, recommendations can be proposed to enhance development, making it more accessible for villages to be electrified from PLN or to increase the electrification of villages currently not using electricity. source of electricity from non-PLN to realize the Bright Indonesia Program.