Predictive Modeling with Pipelines in tidymodels and scikit-learn

In this workshop, we will use predictive modeling to analyze data about NHL hockey players between 2006 and 2018. The goal is to learn and practice the standard workflow used for machine learning analysis:

  1. Cleaning the data
  2. Visualizing relationships
  3. Choosing predictor (input) variables
  4. Choosing between competing models
  5. Interpreting model conclusions
  6. Ethics considerations

Four colored puzzle pieces where three of the four have been linked together and the fourth (green) piece is about to be placed with the other pieces.

Accessing Workshop Materials

You can either work locally with the workshop materials (on your own laptop) or remotely (online). To work locally, you will need to have either R or Python installed on your computer, and a variety of packages. If you do not have either of these installed, you can still work online!

Local Files

Click here to download the R Quarto file

Click here to download the Python Quarto file

Remote Google Colab Files

Click here for Python notebook

Click here for R notebook