/end2end_ds_workshop

End-to-end data science workshop

Primary LanguageJupyter Notebook

end2end Data Science Workshop

This repository covers the notebooks required for the click prediction workshop.

  1. Exploration - Statistical significance, and Categorical variable ranking with mutual information
  2. Feature Engineering - Time feature engineering, timezone correction, and feature interactions and visualisation.
  3. Data enrichment - Working with scraped files, archives, feature selection.
  4. Modeling - Training, Time-dependent split, model explainability and regularization.
  5. Coefficient Monitoring - Tracking model performance and structure through time.

For more details, contact me at goren.ml .