Machine Learning Algorithms from Start to Finish in Python: SVM

Support Vector Machines are very versatile Machine Learning algorithms. The main reason for their popularity is for their ability to perform both linear and non-linear classification and regression using what is known as the kernel trick; if you don't know what that is, don't worry. So, without further ado, let's dive in! You may be wondering why you need to have another one in your toolkit! Here we see three potential decision boundaries for classifying the data: H1, H2, and H3. First off, H1 does not separate the classes at all, so it is not a good hyperplane.