Electricity Usage Clustering with K-means++ Algorithm

摘要

Global electricity consumption is increasing every year. Electricity has become a necessary need for every sector, from a household to government and industry. With today's technology, data also has become more important, including electricity data. The right tools and techniques can extract valuable information from data. Using a K-Means++ algorithm to cluster electricity data can help to determine when the usage is low, moderate, and high. In this study, there three scenarios of clustering; hourly, daily, and monthly. The silhouette score of this experiment ranges from 0.68 to 0.71, and the DB Index ranges from 0.30 to 0.51.