Python Machine Learning By Example

Scenario 2 - determining the optimal hyperplane

In the following instance, hyperplane C is the optimal one that enables the maximal sum of the distance between the nearest data point in the positive side to itself and the distance between the nearest data point in the negative side to itself. The nearest point(s) in the positive side can constitute a hyperplane parallel to the decision hyperplane, which we call a positive hyperplane; on the other hand, the nearest point(s) in the negative side compose the negative hyperplane. The perpendicular distance between the positive and negative hyperplanes is called margin, whose value equates to the sum of two distances aforementioned. A decision hyperplane is optimal if the margin is maximized. ...