Mondrian: how to use prior knowledge on groups when measuring uncertainty?

• 1. Create the noisy dataset
• 2. Split the dataset into a training set, a conformalization set, and a test set
• 3. Fit a random forest regressor on the training set
• 4. Build the classical conformal prediction intervals
  • Conformalize a SplitConformalRegressor on the conformalization set
  • Predict the prediction intervals on the test set
  • Evaluate the coverage score by group
• 5. Build the Mondrian conformal prediction intervals
  • Conformalize a SplitConformalRegressor on the conformalization set for each group
  • Predict the prediction intervals on the test set
• 6. Compare the coverage by partition, plot both methods side by side
• Theoretical Description
  • References