# An Exact Solver for the Weston-Watkins SVM Subproblem

arXiv: https://arxiv.org/abs/2102.05640 · HTML: https://mlsanity.com/paper/2102.05640/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 6 of its 49 references that are in the map.
- First posted on arXiv 10 Feb 2021.

## Builds on (most cited first, 6 of 6)

- [It's Not Just Size That Matters: Small Language Models Are Also Few-Shot Learners](https://mlsanity.com/paper/2009.07118/) (2020)
- [Self-Normalizing Neural Networks](https://mlsanity.com/paper/1706.02515/) (2017)
- [liquidSVM: A Fast and Versatile SVM package](https://mlsanity.com/paper/1702.06899/) (2017)
- [Efficient Primal-Dual Algorithms for Large-Scale Multiclass Classification](https://mlsanity.com/paper/1902.03755/) (2019)
- [Distributed Block-diagonal Approximation Methods for Regularized Empirical Risk Minimization](https://mlsanity.com/paper/1709.03043/) (2017)
- [Faster SVM Training via Conjugate SMO](https://mlsanity.com/paper/2003.08719/) (2020)

## Abstract (arXiv)

Recent empirical evidence suggests that the Weston-Watkins support vector machine is among the best performing multiclass extensions of the binary SVM. Current state-of-the-art solvers repeatedly solve a particular subproblem approximately using an iterative strategy. In this work, we propose an algorithm that solves the subproblem exactly using a novel reparametrization of the Weston-Watkins dual problem. For linear WW-SVMs, our solver shows significant speed-up over the state-of-the-art solver when the number of classes is large. Our exact subproblem solver also allows us to prove linear convergence of the overall solver.