17.1 Introduction to Support Vector Machines

A Support Vector Machine (SVM) is a supervised learning algorithm that can be used for classification or regression analysis. What makes Support Vector Machines stand out is the fact that these algorithms can classify linear and nonlinear data.

Looking at Figure x.x. below, there are other algorithms that classify data points in space by using different methods. For example, decision trees will use various splits, while linear functions will draw only a straight line to separate the data into different categories. In some situations, these algorithms may perform well at classifying data, while at other times they won’t. Support Vector Machines, however, have the capability of handling all these different types of spaces and its data points. Especially nonlinear functions, which many other algorithms are not capable of doing.

You may wonder how Support Vector Machines can be applied and used in real-life applications. Below are a few examples of how SVMs can be applied:

• Learning to recognize fraudulent credit card activity

• Learning to recognize handwritten digits and letters

• Classification of images

• Automatic classification of microarray gene expression profiles

• Disease identification

• Many biological and other sciences