Python Idle
A house's lot area is measured in tens of thousands of square feet. Its overall quality is rated 1 to 10. When features live on wildly different scales, distance-based models like K-Nearest Neighbors get confused — the big numbers dominate and the small ones barely matter. Scaling fixes this, but not all scalers are equal. StandardScaler assumes a bell curve, MinMaxScaler squishes everything to [0, 1], and RobustScaler shrugs off outliers. In this lab you'll apply all three, visualize the distributions, and see which one helps KNN the most.
~25 minsandbox lab510 rows
Loading Python runtime...
Goals: 5 tests
standard_score should exist as a float
minmax_score should exist as a float
robust_score should exist as a float
scores should not all be identical
matplotlib should have been called
Python loading...