Features & Encoding the Real World

Classical models built from the math you already know

Real data rarely arrives as a clean grid of numbers. A dataset of customers might have a column for city, a column for how someone heard about the product, and a column for household income, and any of those columns can also have gaps where nobody answered. Every model in this course, from a straight line to a decision tree, secretly assumes its input is already a list of numbers with nothing missing. Feature encoding is the work that happens before training even starts, turning categories and blanks into numbers a model can actually use.

🔒 This is a Pro lesson — the interactive figure, worked examples, quiz and practice open with Pro access.

▶ Features & Encoding the Real World
← The Overfitting Dial: Capacity, Train vs Held-outPolynomial & Basis-Expansion Regression →