Support Vector Machines find the widest possible gap between two classes. Click the canvas to add points, adjust the regularization parameter C, and watch the decision boundary adapt.
Low C → wider margin, more misclassifications allowed.
High C → narrow margin, fewer violations tolerated.
The support vectors are the only points that matter for defining the boundary: all others could be removed and the hyperplane wouldn't change. This makes SVMs robust to outliers far from the boundary.