Bracketology with Google Machine Learning

In this lab you use Machine Learning (ML) to analyze the public NCAA dataset and predict NCAA tournament brackets.

• GSP461
• Overview
• Setup and requirements
• Task 1. Open the BigQuery console
• Task 2. NCAA March Madness
• Task 3. Find the NCAA public dataset in BigQuery
• Task 4. Write a query to determine available seasons and games
• Task 5. Understand machine learning features and labels
• Task 6. Create a labeled machine learning dataset
• Task 7. Create a machine learning model to predict the winner based on seed and team name
• Task 8. Evaluate model performance
• Task 9. Making predictions
• Task 10. How many did our model get right for the 2018 NCAA tournament?
• Task 11. Models can only take you so far...
• Task 12. Using skillful ML model features
• Task 13. Preview the new features
• Task 14. Interpreting selected metrics
• Task 15. Train the new model
• Task 16. Evaluate the new model's performance
• Task 17. Inspect what the model learned
• Task 18. Prediction time!
• Task 19. Prediction analysis:
• Task 20. Where were the upsets in March 2018?
• Task 21. Comparing model performance
• Task 22. Predicting for the 2019 March Madness tournament
• Congratulations!