What is feature bagging in Random Forest?

Updated May 16, 2026

Short answer

Feature bagging randomly selects subsets of features at each split to increase model diversity.

Deep explanation

Feature bagging is a key mechanism in Random Forest where each decision tree considers only a random subset of features when splitting nodes. This prevents dominant features from being used across all trees, reducing correlation between trees and improving ensemble performance. It enhances generalization by encouraging different trees to learn different patterns.

Real-world example

Used in medical diagnosis systems to avoid over-reliance on a single biomarker.

Common mistakes

  • Using all features at every split defeats ensemble diversity.

Follow-up questions

  • Why is sqrt used for classification?
  • What happens if max_features = all?

More Ensemble Learning interview questions

View all →