Introduction
How do you know if a student really understands a subject? You don't give them the exact questions they practised — you give them an exam with new questions. Only then do you see whether they learned the method or just memorised answers.
Machine learning works the same way. A model's score on the data it was trained on tells you very little; it might have simply memorised them. To know how it will do in the real world, you must test it on data it has never seen. This topic explains how to split your data so you can do exactly that.