Predicting-Energy-Consumption-in-a-Building-Using-Machine-Learning-Techniques

In this project, I implemented Linear Regression, Ridge Regression, Lasso Regression, KNN Regression , Decision Tree Regression, Random Forest Regression, Gradient Boosting Regression, and Artificial Neural Network Regression with Python on Kaggle “Energy Consumption” dataset to forecast energy consumption of a house. First of all, I explored and then visualized data using some plots. After that, the machine learning project starts; 8 aforementioned regression models are implemented and each model is evaluated using 4 functions: MSE, MAE, MAPE, and R-square. Comparing the R-squares of each model, it is noticeable that Random Forest Regression has the highest R-square (0.9). After that, I tried to improve the accuracy of models with 2 techniques, Feature Selection and Feature Scaling. After recruiting these techniques, the linear model with R-square = 0.99 has the highest accuracy.