Faster Coordinate Descent via Adaptive Importance Sampling
Abstract
Coordinate descent methods employ random partial updates of decision variables in order to solve huge-scale convex optimization problems. In this work, we introduce new adaptive rules for the random selection of their updates. By adaptive, we mean that our selection rules are based on the dual residual or the primal-dual gap estimates and can change at each iteration. We theoretically characterize the performance of our selection rules and demonstrate improvements over the state-of-the-art, and extend our theory and algorithms to general convex objectives. Numerical evidence with hinge-loss support vector machines and Lasso confirm that the practice follows the theory.
Cite
Text
Perekrestenko et al. "Faster Coordinate Descent via Adaptive Importance Sampling." International Conference on Artificial Intelligence and Statistics, 2017.Markdown
[Perekrestenko et al. "Faster Coordinate Descent via Adaptive Importance Sampling." International Conference on Artificial Intelligence and Statistics, 2017.](https://mlanthology.org/aistats/2017/perekrestenko2017aistats-faster/)BibTeX
@inproceedings{perekrestenko2017aistats-faster,
title = {{Faster Coordinate Descent via Adaptive Importance Sampling}},
author = {Perekrestenko, Dmytro and Cevher, Volkan and Jaggi, Martin},
booktitle = {International Conference on Artificial Intelligence and Statistics},
year = {2017},
pages = {869-877},
url = {https://mlanthology.org/aistats/2017/perekrestenko2017aistats-faster/}
}