juniorDimensionality Reduction
Difference between feature selection and feature extraction
Updated May 16, 2026
Short answer
Feature selection chooses existing features; feature extraction creates new transformed features.
Deep explanation
Feature selection removes irrelevant or redundant features, while feature extraction transforms data into a lower-dimensional space like PCA. Extraction often improves representation power.
Real-world example
Selecting important genes vs combining gene expressions into latent variables.
Common mistakes
- Confusing PCA with feature selection.
Follow-up questions
- When is feature selection preferred?
- When is feature extraction preferred?