Dual SVM Training on a Budget

Sahar Qaadan, Merlin Schüler, Tobias Glasmachers

Abstract

We present a dual subspace ascent algorithm for support vector machine training that respects a budget constraint limiting the number of support vectors. Budget methods are effective for reducing the training time of kernel SVM while retaining high accuracy. To date, budget training is available only for primal (SGD-based) solvers. Dual subspace ascent methods like sequential minimal optimization are attractive for their good adaptation to the problem structure, their fast convergence rate, and their practical speed. By incorporating a budget constraint into a dual algorithm, our method enjoys the best of both worlds. We demonstrate considerable speed-ups over primal budget training methods.

Download


Paper Citation


in Harvard Style

Qaadan S., Schüler M. and Glasmachers T. (2019). Dual SVM Training on a Budget.In Proceedings of the 8th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-351-3, pages 94-106. DOI: 10.5220/0007346400940106


in Bibtex Style

@conference{icpram19,
author={Sahar Qaadan and Merlin Schüler and Tobias Glasmachers},
title={Dual SVM Training on a Budget},
booktitle={Proceedings of the 8th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2019},
pages={94-106},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0007346400940106},
isbn={978-989-758-351-3},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 8th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - Dual SVM Training on a Budget
SN - 978-989-758-351-3
AU - Qaadan S.
AU - Schüler M.
AU - Glasmachers T.
PY - 2019
SP - 94
EP - 106
DO - 10.5220/0007346400940106