- Exploratory Data Analysis & Visualization
- Dealt with missing values
- Log target variables for better distribution
- Normalized data using skew and Box Cox transformations
- Applied RobustScaler on number columns
- Shifted outliers of prime attributes closer to the linear regression and used ZScore to drop outliers of other attributes.
- experimented with: xgBoost, Lasso, Ridge, SVR, LGBM, Gradient Boosting Regressor, Keras, Tensor Flow, ENet, KRR.