A stochastic coordinate descent primal-dual algorithm with dynamic stepsize for large-scale composite optimization
Optimization and Control
2016-04-15 v1
Abstract
In this paper we consider the problem of finding the minimizations of the sum of two convex functions and the composition of another convex function with a continuous linear operator. With the idea of coordinate descent, we design a stochastic coordinate descent primal-dual splitting algorithm with dynamic stepsize. Based on randomized Modified Krasnosel'skii-Mann iterations and the firmly nonexpansive properties of the proximity operator, we achieve the convergence of the proposed algorithms. Moreover, we give two applications of our method.
Cite
@article{arxiv.1604.04172,
title = {A stochastic coordinate descent primal-dual algorithm with dynamic stepsize for large-scale composite optimization},
author = {Meng Wen and Shigang Yue and Yuchao Tang and Jigen Peng},
journal= {arXiv preprint arXiv:1604.04172},
year = {2016}
}
Comments
arXiv admin note: substantial text overlap with arXiv:1407.0898 by other authors