Build your March Madness bracket with machine learning!

The app has 4 steps:

1. Parameter Selection (this page) - Choose the stat distributions for model training.
2. Model Selection - Pick your ML model and tune hyperparameters.
3. Testing Methods - Configure advanced training options (ensemble, round splitting, etc).
4. Results - View your predicted bracket with scoring.

Two ways to pick parameters:

• Quick Select: Use the slider to pick N parameters automatically. Uses feature selection to find the least-correlated stats.
• Manual Select: Browse categories and hand-pick individual stats. Each stat has Normal (_N), Difference (_D), and Smart Difference (_SD) variants.

If unsure, Quick Select with ~30-50 parameters is a great starting point!