Kaggle-Competition-Sberbank

Top 1% rankings (22/3270) code sharing for Kaggle competition Sberbank Russian Housing Market: https://www.kaggle.com/c/sberbank-russian-housing-market

• Data.py: Data cleaning and feature engineering
• Exploration.py: Used for unorganized data exploration
• Model.py: Main model (xgboost)
• BaseModel.py: Base models for ensembling
• lightGBM.py: lightgbm model for ensembling
• Stacking.py: model stacking (final model)

Final pipeline: Raw data → Data.py → Stacking.py.
Model.py, BaseModel.py and lightGBM.py are not used for the final output.
Original data set can be downloaded from the competition page.