Introduction and Developments on Support Vector Machines and Classification



Support vector machines (SVM) have become a valuable method for the prediction of two classes. Often, the SVM is the best method to classify the two groups. One of the aims of this workshop is to introduce the technique of support vector machines. Another aim is to discuss the use of kernels to allow for nonlinearity of the predictor variables. This workshop brings together several leading researchers active this area in the Netherlands and two foreign guests, Sarel Steel (Stellenbosch University, South Africa) and Thorsten Joachims (Cornell University NY, USA). 

10:00 - 10:30 Welcome with coffee
10:30 - 11:10 Ida Sprinkhuizen-Kuyper (Radboud University Nijmegen):
  Introduction to Support Vector Machines, a powerful tool for classification and prediction
11:10 - 11:50 Patrick Groenen (Erasmus University Rotterdam):
  Support Vector Machines using Majorization and Kernels
11:50 - 12:00 Coffee
12:00 - 12:40 Evgueni N. Smirnov (Maastricht University):
  Version Space Support Vector Machines
12:40 - 13:45 Lunch
13:45 - 14:25 Georgi Nalbantov (Maastricht University):
  Instance-based Penalization Methods for Classification
14:25 - 15:05 Sarel Steel (Stellenbosch University, South Africa):
  Variable selection for kernel methods
15:05 - 15:30 Break
15:30 - 16:20 Thorsten Joachims (Cornell University NY, USA):
  Support Vector Machines for Structured Output Prediction
16.20 Drinks
Participation in the workshop is free. To know the number of participants in advance, please register by sending an e-mail to Elli Hoek van Dijke (
For a full program with abstracts and directions, see here.
Contact information:
Patrick Groenen