Support Vector Machines (SVM) is an important and widely used machine learning algorithm. In order to fully understand SVMs, you need to have a fundamental understanding of how the statistical learning method functions. Here is a useful lecture on SVM coming from MIT OpenCourseware. In the lecture, Professor Patrick H. Winston explores support vector machines in some mathematical detail. He uses Lagrange multipliers to maximize the width of the street given certain constraints. He also shows how to transform vectors into another space, using a kernel function.
The lecture comes from course MIT 6.034 Artificial Intelligence, Fall 2010. You can view the complete course: http://ocw.mit.edu/6-034F10
Earn your master’s in predictive analytics completely online from Northwestern University.