Interview OS
Back to AI / ML

Library

Feature engineering

Encoding, scaling, leakage.

Data

Overview

Feature engineering transforms raw data into signals a model can learn from — encoding, scaling, interactions, and domain-derived features. It often moves the needle more than swapping algorithms.

How it works

Data
RawTransformSelectOutputDatanulls/typesFeaturesstandardizeImportancecorr/SHAPFeature set
ClientServiceEdgeData

Step by step, with examples

  1. 1

    Data

    • Clean, impute, and encode.
  2. 2

    Features

    • Scale, one-hot, and add interactions.
  3. 3

    Importance

    • Drop noise and prevent leakage.
  4. 4

    Feature set

    • Feed clean features to the model.
    • Example: no target leakage

Overview

Transform raw data into informative features: encoding categoricals, scaling, interactions, and time-aware splits.

Common pitfalls

  • Target leakage from future data
  • Fitting scalers on the full dataset before split

Where this content comes from

For full transparency, this content is curated and verified from these sources:

Peer-reviewed ML papers & model cardsProduction ML engineering referencesOppZen-authored ML eval rubrics