Towards Energy Efficiency Smart Buildings Models Based on Intelligent Data Analytics

This work presents how to proceed during the processing of all available data coming from smart buildings to generate models that predict their energy consumption. For this, we propose a methodology that includes the application of different intelligent data analysis techniques and algorithms that have already been applied successfully in related scenarios, and the selection of the best one depending on the value of the selected metric used for the evaluation. This result depends on the specific characteristics of the target building and the available data. Among the techniques applied to a reference building, Bayesian Regularized Neural Networks and Random Forest are selected because they provide the most accurate predictive results.