Support vector machines and kernels

Many lines can separate two classes, and a support vector machine takes the one with the widest margin. This topic finds that line from the points at its edge, then uses kernels to separate classes that no straight line can.