What is bootstrap sampling?

Updated May 17, 2026

Short answer

Bootstrap sampling is randomly sampling data with replacement to create multiple training datasets.

Deep explanation

Each tree in Random Forest is trained on a bootstrap sample, meaning some rows may repeat while others may be left out. This introduces diversity among trees.

Real-world example

Used in ensemble models for medical diagnosis systems to improve robustness.

Common mistakes

  • Assuming bootstrap samples are unique subsets.

Follow-up questions

  • Why use replacement in sampling?
  • What are out-of-bag samples?

More Random Forest interview questions

View all →