A First-Order Stochastic Primal-Dual Algorithm with Correction Step

被引:10
|
作者
Rosasco, Lorenzo [1 ,2 ,3 ]
Villa, Silvia [4 ]
Bang Cong Vu [1 ,2 ]
机构
[1] Ist Italiano Tecnol, Lab Computat & Stat Learning, Genoa, Italy
[2] MIT, 77 Massachusetts Ave, Cambridge, MA 02139 USA
[3] Univ Genoa, Dept Informat Bioengn Robot & Syst Engn DIBRIS, Genoa, Italy
[4] Politecn Milan, Dipartimento Matemat, Milan, Italy
关键词
Cocoercive operator; composite operator; duality; monotone inclusion; maximal monotone operator; operator splitting; primal-dual algorithm; stochastic errors; SADDLE-POINT PROBLEMS; MONOTONE INCLUSIONS; SPLITTING ALGORITHM; PROXIMAL METHODS; CONVERGENCE; SPARSITY; REGULARIZATION; OPTIMIZATION; OPERATORS;
D O I
10.1080/01630563.2016.1254243
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
In this article, we investigate the convergence properties of a stochastic primal-dual splitting algorithm for solving structured monotone inclusions involving the sum of a cocoercive operator and a composite monotone operator. The proposed method is the stochastic extension to monotone inclusions of a proximal method studied in the literature for saddle point problems. It consists in a forward step determined by the stochastic evaluation of the cocoercive operator, a backward step in the dual variables involving the resolvent of the monotone operator, and an additional forward step using the stochastic evaluation of the cocoercive operator introduced in the first step. We prove weak almost sure convergence of the iterates by showing that the primal-dual sequence generated by the method is stochastic quasi-Fejer-monotone with respect to the set of zeros of the considered primal and dual inclusions. Additional results on ergodic convergence in expectation are considered for the special case of saddle point models.
引用
收藏
页码:602 / 626
页数:25
相关论文
共 50 条
  • [21] First-order primal-dual algorithm for image restoration corrupted by mixed Poisson-Gaussian noise
    Chen, Miao
    Wen, Meng
    Tang, Yuchao
    SIGNAL PROCESSING-IMAGE COMMUNICATION, 2023, 117
  • [22] A first-order inexact primal-dual algorithm for a class of convex-concave saddle point problems
    Jiang, Fan
    Wu, Zhongming
    Cai, Xingju
    Zhang, Hongchao
    NUMERICAL ALGORITHMS, 2021, 88 (03) : 1109 - 1136
  • [23] A General Framework of Exact Primal-Dual First-Order Algorithms for Distributed Optimization
    Mansoori, Fatemeh
    Wei, Ermin
    2019 IEEE 58TH CONFERENCE ON DECISION AND CONTROL (CDC), 2019, : 6386 - 6391
  • [24] Bregman primal-dual first-order method and application to sparse semidefinite programming
    Jiang, Xin
    Vandenberghe, Lieven
    COMPUTATIONAL OPTIMIZATION AND APPLICATIONS, 2022, 81 (01) : 127 - 159
  • [25] ACCELERATION AND GLOBAL CONVERGENCE OF A FIRST-ORDER PRIMAL-DUAL METHOD FOR NONCONVEX PROBLEMS
    Clason, Christian
    Mazurenko, Stanislav
    Valkonen, Tuomo
    SIAM JOURNAL ON OPTIMIZATION, 2019, 29 (01) : 933 - 963
  • [26] FlexPD: A Flexible Framework of First-Order Primal-Dual Algorithms for Distributed Optimization
    Mansoori, Fatemeh
    Wei, Ermin
    IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2021, 69 : 3500 - 3512
  • [27] Exploiting Strong Convexity from Data with Primal-Dual First-Order Algorithms
    Wang, Jialei
    Xiao, Lin
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 70, 2017, 70
  • [28] A Preconditioning Technique for First-Order Primal-Dual Splitting Method in Convex Optimization
    Wen, Meng
    Peng, Jigen
    Tang, Yuchao
    Zhu, Chuanxi
    Yue, Shigang
    MATHEMATICAL PROBLEMS IN ENGINEERING, 2017, 2017
  • [29] On the ergodic convergence rates of a first-order primal–dual algorithm
    Antonin Chambolle
    Thomas Pock
    Mathematical Programming, 2016, 159 : 253 - 287
  • [30] A STOCHASTIC COORDINATE DESCENT PRIMAL-DUAL ALGORITHM AND APPLICATIONS
    Bianchi, Pascal
    Hachem, Walid
    Franck, Iutzeler
    2014 IEEE INTERNATIONAL WORKSHOP ON MACHINE LEARNING FOR SIGNAL PROCESSING (MLSP), 2014,