Support vector machine (SVN)
A very simple explanation
A Support Vector Machine (SVM) is one of the widely used algorithms in Machine Learning. In the simple implementation, it looks similar to the linear regression but can be more precise in more complex classification tasks.
How it works?
From first sight, it looks similar to linear regression, but there are also a few key differences. In the simple case of classification, it also tries to separate two classes by the straight line.
This line is called hyperplane and the method to find it is to find the maximum sum of the distance between closest points from both classes. The line between these points is called the support vector. The distance between support vector and hyperplane is called margin. The more margin — the better, we can separate 2 classes more confidently.
Regularisation parameter
Regularisation parameter, or C — parameter, which affects the so-called decision boundary and outliers tolerance. if C is too large algorithm will try to build a hyperplane between all outliers, which means low bias, but high variance. On the other hand, If C too…
