November 14th, Wednesday 14:15, Room 303, Jacobs Building

Title: A (very) sort introduction to support vector machines

Lecturer: Yair Goldberg

Affiliation : Department of Statistics, University of Haifa

Support vector machines (SVM) are a collection of algorithms for classification and regression which are applicable under mild assumptions. In this talk, I will introduce the main concepts of SVM: loss and risk functions, empirical risk minimization, and reproducing kernel Hilbert spaces. I will also discuss some theoretical results regarding the performance of SVMs.