Permutation Importance: How Much Worse the Model Gets When a Column Is Shuffled
This lesson is part of the course
In this lesson
You measure the toy rent model's error on ten test flats, replace one feature at a time with a shuffled copy, measure the error again and rank the features by how much the model relies on them.
Buy the course once for $9.99 to unlock every lesson and lab, the final exam and the certificate – lifetime access, no subscription.
Lessons in this course
- Baseline & Contributions: Explaining a Linear Score Feature by Feature Locked
- Coalitions: All Eight Feature Subsets & the Model Output for Each Locked
- Shapley Values: Marginal Contributions over Every Order & the Weighted Formula Locked
- Interactions: How Shapley Splits an Effect Two Features Create Together Locked
- Permutation Importance: How Much Worse the Model Gets When a Column Is Shuffled Locked
- Partial Dependence & ICE Curves: One Feature Across the Whole Fleet Locked
- Counterfactuals: The Smallest Change That Flips a Decision Locked
- The Explanation Report: Global vs Local, Stability & Limitations Locked