Machine learning model for green building design prediction

Abstract

Green building (GB) is a design concept that implements sustainable processes and green technologies in the building's life cycle. However, the design process of GB tends to take longer than conventional buildings due to the integration of various green requirements and performances into the building design. Advanced artificial intelligence (AI) methods such as machine learning (ML) are widely used to help designers do their jobs faster and more accurately. Therefore, this study aims to develop a GB design predictive model utilizing ML techniques that consider four GB design criteria: energy efficiency, indoor environmental quality, water efficiency, and site planning. A dataset of GB projects collected from a private construction company based in Jakarta was used to train and test the ML model. The accuracy of the models was evaluated using mean square error (MSE). The comparison of MSE values of the conducted experiments showed that the combination of the artificial neural network (ANN) method with the IF-ELSE algorithm created the most accurate ML model for GB design prediction with an MSE of 1.3.