| 성과구성 |
Abstract
In these days, the development of predictive models for energy consumption in buildings has been actively studied. Among various energy prediction methods, unlike existing tools such as simulation, machine learning which has great advantage in that it does not require detailed information of building system is attracting much attention. In this study, the authors develop data-driven machine learning models and identify whether this can be positively applied to predict energy consumption. In addition, we verify and compare the predictive performance of models developed based on deep neural network (DNN), random forest (RF), and gradient boosting machine(GBM). |
