Lecture 6 (Part 1) - Support Vector Machine (SVM) and Kernel Trick - Machine Learning Course
Machine Learning course taught by Dr. Mohamed-Rafik Bouguelia, at Halmstad University, Sweden.
This lecture presents Support Vector Machines (SVM). It starts by presenting the intuition behind the SVM classifier. Then, it explains the optimization problem for a linear SVM with a hard margin and shows how to solve it. Then, it presents SVM with a soft margin and explains the cost function, the hinge loss, and how generalization and overfitting are controlled. Finally, it explains how to do nonlinear classification with SVM (nonlinear SVM) using a kernel function and the kernel trick.