Train, Test, and Validation Sets
Core route · model intuition
Split the data, drag the examples, watch the boundary react.
This page turns data splitting into a hands-on workflow decision. Move training points, toggle features, and feel why the validation set guides choices while the test set stays untouched. It builds intuition before you open a notebook.
At a glance
Choosing splits is usually taught as a rule of thumb. This makes the mechanics physical: drag data, switch features, and watch how each split changes the model you trust.
Jump to the model scene and move the boundary. Switch features, then scroll into the validation and test sections to see why that interaction belongs in the middle of the workflow rather than at the end.
About 10 to 15 minutes. Beginner friendly. One of the easiest machine-learning routes to start with because the workflow logic is visible immediately.
Embed
Reading path
- Open the live interactive: https://kohnnn.github.io/interactive-explanation/train-test-validation/
- Continue through Principal-Component-Analysis or The-Bias-Variance-Tradeoff when you want the next layer of model intuition
- Continue via Interactive or Visual Notes