🗂️ Classification — sorting with labelled examples
Show the computer lots of examples that are already labelled. It learns a
boundary between the two groups. Then, for something brand-new, it decides which side it's on.
Try all three problems below — same idea, different data!
Pick a problem
🤖 In a real AI — this is “supervised” learning
Every one of these needs a teacher who first labels the examples (healthy / junk,
cat / dog, grow / wilt, spam / not-spam…). The model learns the boundary, then labels new things by
which side they fall on. Real models use many features (not just 2) and can draw curvy
boundaries — but the idea is exactly this.