Check your understanding
Quiz
You have turned raw columns into features a model can use: encoding categories without inventing order, target-encoding without leaking the answer, scaling and reshaping numbers, building interactions and splines, pulling features out of dates and text, imputing what is missing, and selecting the features that earn their place. This quiz checks what stuck. The last two steps are live R you can run.
Question 1 of 6
Encoding a nominal category
You have a city column with values London, Paris and Tokyo. Why is plain integer (label) encoding a poor choice here?
Correct: nominal categories have no order, but 1/2/3 tells a model Tokyo is "three times" London.
Overfitting is not the core issue; the fake ordinal relationship is.