AI Glossary Term
What does “Feature Engineering” mean?
Definition
Designing or transforming input variables to make learning easier and more effective.
How Feature Engineering works
Example
A taxi-demand model receives raw timestamps. An engineer converts them into hour of day, day of week, and holiday flags, which makes patterns such as rush hour easy for the model to pick up.
Why it matters
On table-based problems, thoughtful features often improve results more than switching algorithms. The work needs domain knowledge.
Common questions
Is feature engineering still needed with deep learning?
Less for images and text, where networks learn features. It remains important for structured business data.
What is a leakage risk here?
Building features using information that would not be available at prediction time, which makes test scores look better than real performance.
How do I know a feature helps?
Compare model performance with and without it on held-out data.