This demo will be revisiting the Kaggle Titanic dataset
We'll be introducing some new tools to implement what we did last session. Using custom classes (regressors, classifiers, cluster-ers, transformers, feature unions, and pipelines) can be powerful additions to your tool belt.
This introduction is modeled after Adam Rogers's titanic_finished-ish.py script we worked through last time.
Data available at https://www.kaggle.com/c/titanic/data.