 Support vector machines (SVMs) are a set of related supervised learning methods used for classification and regression.
 The key is the use of the "kernel trick" to map a nonlinear classification problem in lower dimensional input space
to a linear classification problem in higher dimensional feature space.
