Support vectors in SVM
SVM is an algorithm that can produce significantly accurate results with less computation power. It is widely used in data classification tasks. If a dataset has n number of features, SVM finds a hyperplane in the n-dimensional space, which is also called the decision boundary, to classify the data points. An optimal decision boundary maximizes the distance between the boundary and instances in both classes. The distance between data points in the classes (shown in Figure 5.1a) is known as the margin:
Figure 5.1a: Optimal hyperplane
An SVM algorithm finds the optimal line in two dimensions or the optimal hyperplane in more than two dimensions that separates the space into classes. The optimal hyperplane or optimal line maximizes the margin (the distance between the data points of the two classes). In 3D (or more), data points become vectors and those (very small subset of training examples) that are closest to or on the hyperplanes...