Optimization over zonotopes and training support vector machines
M Bern, D Eppstein - Workshop on Algorithms and Data Structures, 2001 - Springer
Workshop on Algorithms and Data Structures, 2001•Springer
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
C-SVMs for C→∞.
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
C-SVMs for C→∞.
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 C-SVMs for C → ∞.
Springer