/Quantitative-calculation-of-energy-consumption-of-residential-buildings

Prediction of HL and CL

Primary LanguageJupyter Notebook

Quantitative calculation of energy consumption of residential-buildings

  • Utilizing machine learning methods to estimate the EPB of buildings

Goal

  • Predict Heating Load and Cooling Load

Input Parameters

  • Relative compactness
  • Surface area
  • Wall area
  • Roof area
  • Overall height
  • Orientation
  • Glazing area

Output Parameters

  • Heating Load
  • Cooling Load

Requirements

  • Numpy
  • Pandas
  • Sklearn
  • Matplotlib

Results

  • For Heating Load
Algorithm Accuracy ( STD)
LiR 0.892809 (0.064143)
Ridge 0.888898 (0.073554)
Lasso 0.739750 (0.185156)
ElasticNet 0.751113 (0.184069)
Bag_Re 0.969104 (0.082573)
RandomForest 0.968549 (0.084006)
ExtraTreesRegressor 0.968994 (0.082331)
KNN 0.885004 (0.194983)
CART 0.968717 (0.082255)
SVM 0.837300 (0.149457)
  • For Cooling Load
Algorithm Accuracy ( STD)
LiR 0.876337 (0.033166)
Ridge 0.870432 (0.036123)
Lasso 0.751754 (0.101792)
ElasticNet 0.762353 (0.098235)
Bag_Re 0.963511 (0.023490)
RandomForest 0.963513 (0.024185)
ExtraTreesRegressor 0.950593 (0.022661)
KNN 0.924107 (0.088863)
CART 0.955078 (0.021109)
SVM 0.859163 (0.076793)