4.6 Article

Structured nonconvex and nonsmooth optimization: algorithms and iteration complexity analysis

期刊

COMPUTATIONAL OPTIMIZATION AND APPLICATIONS
卷 72, 期 1, 页码 115-157

出版社

SPRINGER
DOI: 10.1007/s10589-018-0034-y

关键词

Structured nonconvex optimization; E-Stationary; Iteration complexity; Conditional gradient method; Alternating direction method of multipliers; Block coordinate descent method

资金

  1. NSFC [11771269, 11831002]
  2. Program for Innovative Research Team of Shanghai University of Finance and Economics
  3. National Science Foundation [CMMI-1462408]
  4. Shenzhen Fundamental Research Fund [KQTD2015033114415450]

向作者/读者索取更多资源

Nonconvex and nonsmooth optimization problems are frequently encountered in much of statistics, business, science and engineering, but they are not yet widely recognized as a technology in the sense of scalability. A reason for this relatively low degree of popularity is the lack of a well developed system of theory and algorithms to support the applications, as is the case for its convex counterpart. This paper aims to take one step in the direction of disciplined nonconvex and nonsmooth optimization. In particular, we consider in this paper some constrained nonconvex optimization models in block decision variables, with or without coupled affine constraints. In the absence of coupled constraints, we show a sublinear rate of convergence to an E-stationary solution in the form of variational inequality for a generalized conditional gradient method, where the convergence rate is dependent on the Holderian continuity of the gradient of the smooth part of the objective. For the model with coupled affine constraints, we introduce corresponding E-stationarity conditions, and apply two proximal-type variants of the ADMM to solve such a model, assuming the proximal ADMM updates can be implemented for all the block variables except for the last block, for which either a gradient step or a majorization-minimization step is implemented. We show an iteration complexity bound of O(1/E2) to reach an E-stationary solution for both algorithms. Moreover, we show that the same iteration complexity of a proximal BCD method follows immediately. Numerical results are provided to illustrate the efficacy of the proposed algorithms for tensor robust PCA and tensor sparse PCA 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