Support Vector Machine Overview

This lecture provides an overview of Support Vector Machines (SVM), explaining how SVM maps input space to a higher dimension feature space using a kernel function. It compares SVM to other classifiers, highlighting advantages such as building a model with a global optimum guarantee, but also discussing disadvantages like computational cost and the need to choose hyperparameters. The lecture concludes by discussing when to use SVM and its limitations, such as being limited to two classes and lacking a notion of confidence.