Home▸Machine Learning & Neural Networks▸Perceptron 2D — Decision Boundary Trainer (2D)

🧠 Perceptron 2D — Decision Boundary Trainer (2D)

Interactive 2D perceptron trainer: click to place points of two classes, watch the decision boundary line converge live via the classic Rosenblatt learning rule w += lr*(target-predicted)*x.

Machine Learning & Neural Networks2DEasy60 FPS📱 Mobile-adapted⇄ 3D version

⚙ Under the hood

Click to place points of two linearly separable classes and train a perceptron live: the classic learning rule w += lr*(target-predicted)*x updates the weights point by point while the decision boundary line converges on the canvas. Generate random separable data, adjust the learning rate, and step through epochs or train straight to convergence.

machine-learningperceptronclassifierlinear-algebraneural-network

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install