We introduce Universum learning for multiclass problems and propose a novel formulation for multiclass universum SVM (MU-SVM). We also propose a span bound for MU-SVM that can be used for model selection thereby avoiding resampling. Empirical results demonstrate the effectiveness of MU-SVM and the proposed bound.
@article{arxiv.1609.09162,
title = {Universum Learning for Multiclass SVM},
author = {Sauptik Dhar and Naveen Ramakrishnan and Vladimir Cherkassky and Mohak Shah},
journal= {arXiv preprint arXiv:1609.09162},
year = {2016}
}