Feature engineering

Understands that ML models need numerical inputs; knows what normalisation means.

Applies one-hot encoding, normalisation, log transforms, handles missing values with common strategies.

Creates interaction features, target encoding, time-based features; uses automated FE (Featuretools); builds reproducible pipelines.

Designs domain-specific features for tabular domains (finance, NLP, time series); applies feature selection (SHAP, permutation importance).

Contributes to automated feature generation, neural feature interaction models, or data-centric AI methods.