Go FurtherHow models are trained and tested
A closer look at training, testing and the evidence needed to judge whether a model works well.
What happens during training
Build on the basic explanation of learning patterns from examples.
Why testing uses separate examples
Explain why remembering training examples is not the same as working well in new situations.
Different ways to measure success
Show why accuracy alone may hide important kinds of mistakes.
From a test to the real world
Explain why performance can change when people, conditions or data change.