Technical interview guide

Feature Engineering & Selection

Turning raw data into model-ready inputs, and choosing which ones actually help.

Read: 25 min
Practice MCQs: 25
Interview QA: 25
Edition: v4
Editorial status: Reviewed
Relevant for: AI EngineerMLOps EngineerMachine Learning Engineer

Scope: scikit-learn stable preprocessing, composition, selection, and inspection guidance and Google ML Crash Course feature guidance accessed 2026-08-31..

Interview QA

Treat each question like a live interview question: answer out loud first (structure, assumptions, tradeoffs), then open the model answer to spot gaps and rehearse a tighter follow-up.

Curated: 2026-10-05 · Written: 2026-10-05 · Reviewed: 2026-10-03

QA-1

Design a point-in-time-correct feature pipeline for transaction risk.

QA-2

How would you audit a feature set for leakage?

QA-3

How do you choose transformations for numerical features?

QA-4

How would you handle missing values in a production model?

QA-5

How do you encode a high-cardinality categorical feature?

QA-6

Explain leakage-safe target encoding.

QA-7

When are feature crosses useful, and how do you control them?

QA-8

How do you engineer time-based features safely?

QA-9

Design a leakage-safe text feature pipeline.

QA-10

Compare filter, wrapper, and embedded feature selection.

QA-11

How do you prevent feature selection from overfitting validation?

QA-12

How should feature importance be interpreted?

QA-13

How do correlated features affect selection and explanation?

QA-14

How do you evaluate sensitive attributes and proxy features?

QA-15

How do you guarantee offline-online feature parity?

QA-16

What problem does a feature store solve, and what does it not solve?

QA-17

What should be monitored for production features?

QA-18

How do you detect, isolate, and mitigate training-serving skew in online feature pipelines?

QA-19

How should a feature be retired?

QA-20

How do you backfill a corrected feature without corrupting experiments?

QA-21

Design an entity-level rolling aggregate feature.

QA-22

How do you use ablation studies for feature decisions?

QA-23

How do privacy and latency constraints shape feature engineering?

QA-24

What tests should a feature transformation library include?

QA-25