Support Vector Machines (SVMs) are a class of supervised machine learning algorithms mainly used
for classification problems.
Research on Support Vector Machines started in the early 1960s through work by Vapnik and
Chervonenkis. In their current form, SVMs were introduced by Boser, Cortes, Guyon and Vapnik in the
1990s.
They are today among the most common machine learning classifiers.
In a binary classification problem, we have a space with points that belong to two different categories.
To classify new points into one of these categories, the Support Vector Machine takes as input the set
of labeled points (the training set) and outputs a hyperplane that divides the space into two regions.
In this talk we focus on the geometric aspects of SVM classifiers. We will look at soft margins, reduced
convex hulls and the kernel trick.