Why is it called Support Vector Machine(SVM)?

Support Vector Machine(SVM) is a supervised machine learning technique. It is used for both classification and regression analysis. The purpose of the support vector machine is to find a hyperplane in an N-dimensional space that can classify the data points.

The shape of the hyperplane depends on the number of features, which are also known as vectors. If there are only two feature vectors, then the hyperplane will be a line only. If there are three feature vectors, the hyperplane turns two-dimensional for the data classification.

Similarly, if the feature vectors are four, then the hyperplane turns to be three-dimensional. Imagine there are more data points. Therefore, the more feature vectors and the more dimensional hyperplane is formed for the data classification.

Let us consider an example as shown in the photo above. We have taken two features, and here we can see the hyperplane is just a line. However, we can see in the figure that we can draw multiple lines to segregate the data points for classification. So, the question arises how do we choose the best hyperplane or the line in the above case for the classification.

What happens when the data points are a bit mixed? 

As we can see in the above case that a blue dot is mixed with the red dots. In such a case the SVM treats the blue dot or the data point as an outlier and does the classification effectively. However, it finds a soft margin distance between the points penalizing each outlier data point with a hinge loss. SVM applies the kernel method to classify the non-linear data points. 

Advantages of using SVM:

1. Robust to deal with outliers 
2. Good in high-dimensional cases 
3. It is memory efficient as it uses training points in decision functions called support vectors. 

Okay, so now you would have been thinking that I still haven't got my answer why is it called a support vector machine. And the answer is that to find the hard and soft margin distances between the points for the hyperplane the initial vectors are used as support. This is why it is called a support vector machine.