Comparative Analysis of Machine Learning Algorithms for the Building Energy Prediction

The construction industry consumes 35% of all global energy. Building energy conservation is critical for lowering emissions and consumption. Properly functioning the building's heating, ventilation, and air conditioning (HVAC) unit helps to reduce energy consumption. Predicting building energy consumption with machine learning (ML) models can help to improve HVAC functionality. As a result, the performance of various ML predictive models based on k-nearest neighbor (KNN), artificial neural network (ANN), support vector regression (SVR), and Ridge and Lasso regression models is investigated in this work for the prediction of energy usage. Furthermore, Bayesian optimization for different random states (RS) is used to estimate the hyperparameters of the ML models that have been implemented. The results show that ANN performs best for RS values between 0 and 75. However, SVR achieves the lowest RMSE for RS, equal to 25, 50, 100, 150, and 200, compared to ANN, KNN, Ridge, and Lasso (RMSE=2.910), respectively. Finally, SVR predicts energy consumption more accurately than other designed models in most cases.