Learning from examples

A model starts out knowing nothing useful. It is shown examples (an email and whether it was spam, a photo and its label, a sentence and the word that came next) and nudges its internal numbers, called weights or parameters, so its guesses get closer to the right answers. Repeat that millions of times and patterns emerge that nobody wrote down.

Why it matters: Training a neural network is exactly this process, carried out with calculus.

Builds on: What AI is and isn't

Learn it, then prove it

Sign in (free) to take this step's mastery check and save your journey.