English

Optimization Over Zonotopes and Training Support Vector Machines

Computational Geometry 2007-05-23 v1 Artificial Intelligence

Abstract

We make a connection between classical polytopes called zonotopes and Support Vector Machine (SVM) classifiers. We combine this connection with the ellipsoid method to give some new theoretical results on training SVMs. We also describe some special properties of soft margin C-SVMs as parameter C goes to infinity.

Keywords

Cite

@article{arxiv.cs/0105017,
  title  = {Optimization Over Zonotopes and Training Support Vector Machines},
  author = {Marshall Bern and David Eppstein},
  journal= {arXiv preprint arXiv:cs/0105017},
  year   = {2007}
}

Comments

To appear in Workshop on Algorithms and Data Structures, 2001