Above shows the learning behavior of a perceptron.

The four input patterns are (1, 1), (1, -1), (-1, -1), and (-1, 1).

Their class are randomly generated and marked by black soft dots (class 1) and black solid dots (class -1).

The yellow region is the solution space if exists.

The green dot marks the initial weight position on the input space and is also randomly generated.

You can reload this page to see different learning behaviors.