A training approach where the model learns from labeled examples (input/output pairs).
Friendly Description: Supervised learning is when an AI learns from examples that come with the right answers attached. It's like learning vocabulary with flashcards: each card shows the word and the definition, so you can check yourself. The AI looks at thousands or millions of these labeled pairs and gradually learns to predict the answer for new examples it has never seen.
Example: To build an AI that detects skin conditions, doctors might prepare thousands of photos, each labeled with the correct diagnosis. The model studies these supervised examples and eventually learns to suggest a likely diagnosis for new photos, all from the patterns it picked up in the labeled training set.