In this lesson
- It starts with a question. A model scores 100% on the photos it was trained on. Is it ready to use?
- The idea you will find. Overfitting is memory pretending to be understanding. — A model that fits the training data too closely learns its accidents — the blue table, the particular camera. The gap between the training score and the held-back score is the measure of it, and closing that gap usually means more varied data rather than more clever maths.
- How you learn it. By doing: sorting.
- Length. 5 short steps. Progress saves on this phone.
Part of Inside AI Models
What a model actually learns from data, where it is strong, and where it fails.