Application of Support Vector Machine and Gaussian Process Regression for Carbon Emission Prediction in Building Construction

Abstract

In light of the heightened awareness of climate change, the construction industry is under significant pressure to reduce its carbon footprint. This study aims to apply two advanced intelligent methods, Support Vector Machine (SVM) and Gaussian Process Regression (GPR), to predict carbon emissions during the building construction stage. The models are trained and tested using four input parameters: quantity of construction machinery, fuel consumption rate, carbon emission factor per unit of fuel or electricity consumed, and operating hours of the machinery. The performance of the models is compared to determine the most accurate and reliable predictor. The results demonstrate that the GPR model consistently outperforms the SVM model in terms of accuracy and consistency. The proposed GPR model is poised to be a valuable tool for policymakers and organizations in making informed decisions to mitigate carbon emissions.