Predictive Models in Python with XGBoost, LightGBM and HyperOpt (Part 1)

In this notebook, I use the following approach to generate predictions on the Kaggle House Price Dataset:

• Data wrangling (Light exploration, followed by removing and transforming some variables)
• Fit a model using a Scikit-Learn pipeline
  • Data Preprocessing + fitting XGBoost/LightGBM estimators with a Randomized Search across their respective hyperparameters
• Evaluate and visualize model performance
• Implement an automated approach to selecting optimal hyperparameters for each estimator (HyperOpt)
• Make predictions