Methods
2.1. Regression algorithm
Relevant independent variables that define the energy consumption are selected to develop models obtained by means of regression models, which are validated and discussed for future predictions and reproducibility.
Take the multiple linear regression for example, suppose variable Y as the power consumption of the building and variable X1, X2 as the impact factors, the general regression formula is as shown as follow :
Y b bX bX (1) 01122
Where b0, b1, b2 are the regression coefficients, where ε is the regression residual.