Some recent developments in the theory of support vector machines

  • Ingo Steinwart (Friedrich-Schiller-Universität Jena Jena, Fakultät für Mathematik und Informatik, Germany)
A3 02 (Seminar room)


In recent years support vector machines (SVM's) became one of the most successful learning algorithms. In this talk I will present an overview of some recent developments in the theory of these algorithms including:
- the role of the kernel and its approximation properties
- asymptotic theory of SVM's
- some lower bounds on the number of support vectors
- incorporating knowledge in nu-SVM's

Finally, I will squetch some interesting open questions.