Support Vector Machines

The shortest distance between the observations and the threshold is called the margin. When we use the threshold that gives the largest margin to make classifications it is call Maximum margin classifier.

  • Eager learners

  • Works for both linear and non-linear classification

  • Discriminative model

  • Leverages max margin classifier

  • Feature scaling

  • Distance based algorithm

Maximum margin classifier #

Sensitive to outliers

Lagrange Duality #

KKT Conditions #

Kernel Tricks #

References #