Accelerated Proximal Stochastic Dual Coordinate Ascent for Regularized Loss Minimization

Abstract

We introduce a proximal version of the stochastic dual coordinate ascent method and show how to accelerate the method using an inner-outer iteration procedure. We analyze the runtime of the framework and obtain rates that improve state-of-the-art results for various key machine learning optimization problems including SVM, logistic regression, ridge regression, Lasso, and multiclass SVM. Experiments validate our theoretical findings.

Cite

Text

Shalev-Shwartz and Zhang. "Accelerated Proximal Stochastic Dual Coordinate Ascent for Regularized Loss Minimization." International Conference on Machine Learning, 2014.

Markdown

[Shalev-Shwartz and Zhang. "Accelerated Proximal Stochastic Dual Coordinate Ascent for Regularized Loss Minimization." International Conference on Machine Learning, 2014.](https://mlanthology.org/icml/2014/shalevshwartz2014icml-accelerated/)

BibTeX

@inproceedings{shalevshwartz2014icml-accelerated,
  title     = {{Accelerated Proximal Stochastic Dual Coordinate Ascent for Regularized Loss Minimization}},
  author    = {Shalev-Shwartz, Shai and Zhang, Tong},
  booktitle = {International Conference on Machine Learning},
  year      = {2014},
  pages     = {64-72},
  volume    = {32},
  url       = {https://mlanthology.org/icml/2014/shalevshwartz2014icml-accelerated/}
}