Cluster Analysis as a Tool for the Territorial Categorization of Energy Consumption in Buildings Based on Weather Patterns

O. May Tzuc; M. Jiménez Torres; Carolina M. Rodriguez; F. N. Demesa López; F. Noh Pat

Resumen:

This book chapter explores the application of k-means, an unsupervised learning technique designed to allow the categorization of patterns and statistical and geographic indicators of energy consumption in various climatic regions of Mexico. It investigates the relationship between energy consumption and climatic and operational patterns in a case study of State Social Housing. The k-means results demonstrate how the distribution of the groups obeys temperature and relative humidity patterns, which can be visualized using Geographic Information Systems software. This methodology has broad implications for future studies on thermal comfort, energy poverty, and pollutant emissions and lays the foundation for replicable research on energy efficiency in housing and other related fields.