Perbandingan XGBoost dan Regresi Linier dalam Mendukung Estimasi Beban Pemanasan dan Beban Pendinginan Bangunan

Abstract

Minimizing the use of cooling and heating devices in buildings can be attempted by considering the building materials used, so that the energy load that will be used can be calculated, as well as saving energy use. In this study, the author chose the XGBoost and linear regression methods to compare and determine the best methods in estimating the heating and cooling loads of \buildings. The author divided the dataset into training data and testing data, 75% as training data and 25% as testing data, using 10-Fold Cross-Validation on training data. In the data processing process, the author uses Jupyter Notebook with the Python programming language. The results of the study using the linear regression method on heating and cooling loads, the author obtained an RMSE value of 3, while the author obtained the smallest RMSE value using the XGBoost method of 1 on heating and cooling loads.