Which set of activities describes feature engineering?
Show answer and explanation
Answer: B
Feature engineering is the work of shaping raw data into inputs that make the underlying pattern easier to learn. It includes choosing relevant columns, encoding categories, scaling numeric ranges, handling missing values, and creating derived quantities such as ratios or time-since-event. Good features can let a simple model outperform a complex one fed raw data. The activity sits in the data preparation stage and depends heavily on domain understanding.
Why the other options are wrong
- A. Hardware procurement is infrastructure planning, not transformation of data into model inputs.
- C. Visualising results happens after modelling and does not change what the model learns from.
- D. Deployment publishes a trained model; it is a later lifecycle stage.
Exam tip. Shaping inputs equals feature engineering.