A perceptron has one obsession: find the cut that best separates the data. Every neural network begins with this deceptively simple move.
It draws a line through data and asks: which side are you on?
In two dimensions, that cut is a line. In three, it becomes a plane. With hundreds of features, it is a hyperplane—hard to picture, but mathematically the same idea.
One perceptron draws one boundary. A neural network combines thousands of them, turning simple cuts into rich, flexible shapes.