Clustering analysis of electricity consumption of municipalities in the province of Pichincha-Ecuador using the k-means algorithm

Abstract

This work shows an analysis of electricity consumption by parishes in the province of Pichincha to define consumption affinities between them. To do this, categorical variables are encoded to be used AQ1 with the unsupervised machine learning algorithm k-means. As relevant results, four types of consumption are established and they are analyzed in relation to population density while algorithm represents 100% of the variability of the data, and municipalities with different consumption trends are identified.