Min-Max scaling vs Z-score standardization: differences and algorithm preferences

What's really being asked

This question probes whether you can move beyond cookbook preprocessing and explain the statistical mechanics behind two fundamental scaling strategies. Interviewers want to see that you understand how the transformed distribution properties interact with an algorithm's objective function, distance metric, or update rule.

The full answer

First, define Min-Max scaling as mapping features to a fixed bounded range, typically zero to one, using the formula X minus X min divided by X max minus X min. It remains sensitive to outliers because the observed minimum and maximum determine the range. Second, define Z-score standardization as centering features at mean zero with standard deviation one using z equals x minus mu divided by sigma, which preserves outlier magnitude and handles varying units. Third, pick an algorithm where one is strongly preferred, such as standardization for PCA because maximizing variance along orthogonal components would otherwise be dominated by variables with larger original scales, or standardization for gradient descent in logistic regression and neural networks because feature values directly modulate weight updates and unequal scales cause some weights to update faster than others.

The mistakes people make

A major red flag is stating that tree-based methods like decision trees or random forests require feature scaling; these algorithms are scale-invariant because splits depend only on rank ordering, not absolute magnitude. Another weak response is claiming Min-Max and Z-score are interchangeable in all contexts without acknowledging that bounded range versus centering at zero matters for optimization stability and outlier sensitivity.

What usually comes next

The interviewer may ask how you would handle a feature with extreme outliers, which usually pushes you toward an outlier-aware method such as robust scaling; neither default Min-Max nor StandardScaler automatically solves extreme values. They might also ask whether you would scale before or after train-test splitting, which tests your awareness of data leakage, or they could ask about the interaction between scaling and regularization in linear models.

A concrete example

Suppose you are building a logistic regression model to predict customer churn from age and annual income. Standardization centers each training feature at zero and scales it by its training standard deviation, which often gives gradient descent a better-conditioned objective. Min-Max scaling instead bounds the observed training range, but an extreme value can stretch that range and place most observations close together. If outliers are substantial, compare an outlier-aware transformation such as RobustScaler. Fit any transformation on training data only, then apply it unchanged to validation and test data.