Statistics for Machine Learning

Support vector machines

Support vector machines are used when the decision boundary is non-linear and would not be separable with support vector classifiers whatever the cost function is! The following diagram explains the non-linearly separable cases for both 1-dimension and 2-dimensions:

It is apparent that we cannot classify using support vector classifiers whatever the cost value is. Hence, we need to use another way of handling the data, called the kernel trick, using the kernel function to work with non-linearly separable data.

In the following diagram, a polynomial kernel with degree 2 has been applied in transforming the data from ...