Data Basics: The Ground Under Every AI

Every AI system is only as good as its data. Good data is correct, complete, current, consistent and representative. Split data into training, validation and test sets, and keep the test set hidden. Check for bias, privacy limits and leakage.

Label

The right answer attached to an example.

Test set

Data kept hidden until the end, for one honest score.

Leakage

When a model sees information it would not have in real use.