Analysis of the Effects of Residential Characteristics on Residential Space Heating Costs in the United States of America with Machine Learning

Abstract

Abstract Energy is one of the main concerns of humanity because energy resources are limited and costly. To reduce the costs and use the energy for residential space heating effectively, it is important to know which factors affect the residential space heating costs. This study aims to analyze the effects of residence characteristics on residential space heating costs in the United States of America by using Bayesian networks, which is a machine learning method. The constructed Bayesian network model shows that the residential space heating costs of the residences are affected mostly by the size of heated residential area. The second most important factor, on the other hand, appears to be major outside wall type, while the third factor is residence type. It is also seen that the insulation levels of the residences seem to have the least effect on the residential space heating costs.