What is one-hot encoding in feature engineering?

Updated May 16, 2026

Short answer

One-hot encoding converts categorical variables into binary vectors.

Deep explanation

It transforms each category into a separate binary column. This prevents machine learning models from interpreting categorical values as ordinal.

Real-world example

Used in customer segmentation based on city or product category.

Common mistakes

  • Creating too many columns leading to high dimensionality.

Follow-up questions

  • What is dummy variable trap?
  • When is label encoding better?

More Feature Engineering interview questions

View all →