ebehlmann / tourney-predictor

Experimenting with machine learning to predict NCAA tournament outcomes

Tourney predictor

This is an experimental project to predict NCAA men's basketball tournament outcomes using a simple machine learning model.

Credits

The data set was provided by Kaggle as part of the Google Cloud & NCAA® ML Competition 2018.

I used a Kaggle-provided Jupyter notebook by Kasper P. Lauritzen as a starting point.

Requirements

To run the Jupyter notebook, the following is required:

• Python
• Jupyter
• Numpy
• Pandas
• Scikit-learn

I recommend downloading and installing Anaconda. Then, you can build a self-contained Python environment that includes all your dependencies.

More information

Read a walkthrough of my process on my blog, Data Storyteller.