Data & Machine Learning

Feature Engineering

8 practice questions. Free questions open a full answer guide; the rest unlock with Pro.

  • A model performed strongly in offline evaluation but its accuracy dropped sharply once deployed. You suspect a feature-engineering issue. How do you investigate whether leakage is the cause, and how do you prevent it going forward? Senior level
  • You want to encode a high-cardinality categorical feature with target encoding to feed a gradient-boosted model. How do you do it without leaking the target, and how would you know if you'd gotten it wrong?Go Pro Senior level
  • What is data leakage in machine learning, and how do you prevent it when scaling or encoding features?Go Pro Junior level
  • You have a categorical feature with thousands of distinct values, like ZIP code or product ID. How would you encode it for a model, and what risks does your choice carry?Go Pro Mid level
  • What is feature engineering, and can you give an example of a useful derived feature?Go Pro Junior level
  • A model that's been stable in production for a year is slowly losing accuracy, and you trace it to a handful of features whose distributions have crept over time. How do you reason about which features to keep, fix, or retire, and how do you keep this from silently happening again?Go Pro Senior level
  • How do you turn a categorical column like country or product-category into something a model can use, and what are the trade-offs between the encodings?Go Pro Junior level
  • You discover mid-sprint that 15% of your training data has null values in a key feature. What do you do?Go Pro Mid level
Want questions matched to your role? Paste a job title, job description, or CV for a personalized set, or go Pro to unlock the full bank.