Composite Objective Mirror Descent
Abstract
We present a new method for regularized convex optimization and analyze it under both online and stochastic optimization settings. In addition to unifying previously known firstorder algorithms, such as the projected gradient method, mirror descent, and forwardbackward splitting, our method yields new analysis and algorithms. We also derive specific instantiations of our method for commonly used regularization functions, such as $\ell_1$, mixed norm, and trace-norm.
Cite
Text
Duchi et al. "Composite Objective Mirror Descent." Annual Conference on Computational Learning Theory, 2010.Markdown
[Duchi et al. "Composite Objective Mirror Descent." Annual Conference on Computational Learning Theory, 2010.](https://mlanthology.org/colt/2010/duchi2010colt-composite/)BibTeX
@inproceedings{duchi2010colt-composite,
title = {{Composite Objective Mirror Descent}},
author = {Duchi, John C. and Shalev-Shwartz, Shai and Singer, Yoram and Tewari, Ambuj},
booktitle = {Annual Conference on Computational Learning Theory},
year = {2010},
pages = {14-26},
url = {https://mlanthology.org/colt/2010/duchi2010colt-composite/}
}