Applied Data Mining-Energy Efficiency of Buildings

Project description

•Used UCI Machine Learning repository data set: https://archive.ics.uci.edu/ml/datasets/Energy+efficiency

• Used Machine learning tools to replace time consuming simulation like Ecotect to accurately predict the Energy Efficiency of Buildings (heating and cooling) .

• Output :Heating Load, Cooling Load

• 8 different parameters: Relative compactness, Surface area, Wall area, Roof Area, Overall Height, Orientation, Glazing Area, Glazing Area Distribution

• Calculated Correlation between different input and output parameters so as to have an idea that how to carry further analysis

• Machine learning tools used : Linear Regression, Regression Tree and Random Forest

• Got Linear Regression model with goodness of fit to be 94.54%

• Regression Tree described 70% samples using input parameters Relative Compactness, Wall Area, Glazing Area

• Random forest gave the idea of importance of input (in descending order) as : Relative compactness, Overall Height, Surface area, Glazing Area, Wall area, Roof Area,Glazing Area Distribution, Orientation