The impact of eight input variables, including relative compactness, roof area, wall area, surface area, glazing area, overall height, orientation, and glazing area distribution, on two output variables, namely, the cooling load (CL) and heating load (HL), of residential buildings has been scrutinized using a statistical machine learning framework. To determine which input variables are most closely associated to each output variable, the association strength between each input variable and each output variable is calculated. To determine which machine learning regression model provides the best accuracy on the dataset and also the variable importance, a comparative analysis based on mean absolute error (MAE), root mean squared error (RMSE), mean squared error (MSE), and R-squared error on different machine learning regression models is conducted.