Clustering Indonesian Provinces Based on Per Capita Energy Consumption Using the K-Means Algorithm

Per capita energy consumption is a crucial indicator for assessing the level of development and energy needs of communities in any region. This research aims to group Indonesian provinces based on their average per capita energy consumption in 2014-2024. The method used is the K-Means Clustering algorithm, an effective data mining technique for identifying hidden patterns and structures in numerical data. The data utilized consists of per capita energy consumption values from each province, which were then processed and analyzed to form several clusters of provinces with similar energy consumption characteristics. The results of the clustering process show that Indonesian provinces can be grouped into three main clusters: provinces with low, medium, and high consumption levels. These findings can serve as a reference for the government in designing more targeted and equitable energy policies, as well as a basis for strategic decision-making in the energy sector and regional development. This study concludes that Indonesia's provincial per capita energy consumption varies significantly across regions and over time, and such variation can be systematically grouped into three distinct clusters—low, medium, and high consumption—using the K-Means Clustering algorithm. The analysis of data from 2018 to 2024, consisting of 238 records across 34 provinces, shows that the majority of provinces (135 out of 238 entries) fall into the medium consumption cluster, with an average of 2,075.85 kcal/capita/day, indicating a relatively stable and balanced energy use that aligns with national standards. In contrast, Cluster 0 represents provinces with lower energy consumption, averaging 1,878.93 kcal, often linked to limited access to modern energy sources and lower economic activity. Meanwhile, Cluster 2, with the highest average of 2,248.75 kcal, comprises industrialized or highly urbanized provinces with more intensive energy needs.