How a Neural Network Makes a Prediction

Using colour as a simple visual example. Unlock each section as you go.

Why this demo exists

This demo intentionally uses a simple colour problem because it is visual and easy to follow. In real software, a fixed colour name can be calculated with rules. Here, the purpose is different: to show how a neural network transforms input values through hidden neurons and arrives at a confidence-based prediction.

Why colour? Colour is easy to see. The real lesson is not colour detection — it is prediction flow.