A WEKA Implementation of

Version Space Support Vector Machines

 

Version Space Support Vector Machines (VSSVMs) are a new approach to reliable classification. They are fully described in:

Smirnov, E.N., Sprinkhuizen-Kuyper, I.G., Nalbantov, G.I., and Vanderlooy, S. (2006). Version Space Support Vector Machines. In: Proceedings of the 17th European Conference on Artificial Intelligence (ECAI 2006), August 28 - September 1, Riva del Garda, Italy, pp.809-810. [PDF]

An extended version of this paper you can find here.  [PDF]

VSSVMs have been implemented as a Java class SeparableVSSVM for the WEKA machine-learning environment. You can download:

Back