FREELessons: 20Length: 2.8 hours

Next lesson playing in 5 seconds

  • Overview
  • Transcript

2.9 Support Vector Machines

Support Vector Machines (SVMs) attempt to find the best decision boundary between two output classes. They are considered the best off-the-shelf classifier because of their excellent performance. In this lesson we give an overview of the math behind SVMs and discuss two modifications, soft margins and kernels, that further improve their performance.