Why does overfitting increase with dimensionality?
Updated May 15, 2026
Short answer
More features allow models to memorize noise.
Deep explanation
High-dimensional feature spaces increase hypothesis space complexity, making it easier to fit training noise.
Real-world example
Spam classifiers overfitting rare token combinations.
Common mistakes
- Using too many irrelevant features.
Follow-up questions
- How to reduce overfitting?
- What is bias-variance tradeoff?