4.6 Article

An optimal method for stochastic composite optimization

期刊

MATHEMATICAL PROGRAMMING
卷 133, 期 1-2, 页码 365-397

出版社

SPRINGER HEIDELBERG
DOI: 10.1007/s10107-010-0434-y

关键词

Stochastic approximation; Convex optimization; Stochastic programming; Complexity; Optimal method; Quadratic penalty method; Large deviation

资金

  1. NSF [CMMI-1000347, CCF-0430644, CCF-0808863]
  2. ONR [N000140811104, N00014-08-1-0033]
  3. Directorate For Engineering
  4. Div Of Civil, Mechanical, & Manufact Inn [1000347] Funding Source: National Science Foundation

向作者/读者索取更多资源

This paper considers an important class of convex programming (CP) problems, namely, the stochastic composite optimization (SCO), whose objective function is given by the summation of general nonsmooth and smooth stochastic components. Since SCO covers non-smooth, smooth and stochastic CP as certain special cases, a valid lower bound on the rate of convergence for solving these problems is known from the classic complexity theory of convex programming. Note however that the optimization algorithms that can achieve this lower bound had never been developed. In this paper, we show that the simple mirror-descent stochastic approximation method exhibits the best-known rate of convergence for solving these problems. Our major contribution is to introduce the accelerated stochastic approximation (AC-SA) algorithm based on Nesterov's optimal method for smooth CP (Nesterov in Doklady AN SSSR 269:543-547, 1983; Nesterov in Math Program 103:127-152, 2005), and show that the AC-SA algorithm can achieve the aforementioned lower bound on the rate of convergence for SCO. To the best of our knowledge, it is also the first universally optimal algorithm in the literature for solving non-smooth, smooth and stochastic CP problems. We illustrate the significant advantages of the AC-SA algorithm over existing methods in the context of solving a special but broad class of stochastic programming problems.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.6
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据