Permutation Feature Importance Lab

A trained classifier's decision only matters as much as the features it actually uses — permutation importance finds out which ones those are without ever opening the model up. This simulator holds 140 points in a 3-feature space where a fixed classification rule labels them, then repeatedly shuffles one feature column at a time across the whole dataset while leaving the true labels and the other two features untouched. Each shuffle is animated directly on the point cloud: points slide along the tested axis to their new shuffled value, and any point that becomes misclassified as a result turns red. Averaging the resulting accuracy drop over K independent shuffles per feature — with the spread reported as a standard deviation — produces the importance bars on the right, exposing that this particular rule leans heavily on two of the three features and almost ignores the third.