ANDERSON ACCELERATED DOUGLAS-RACHFORD SPLITTING

被引:37
|
作者
Fu, Anqi [1 ]
Zhang, Junzi [2 ]
Boyd, Stephen [1 ]
机构
[1] Stanford Univ, Dept Elect Engn, Stanford, CA 94305 USA
[2] Stanford Univ, ICME, Palo Alto, CA 94304 USA
来源
SIAM JOURNAL ON SCIENTIFIC COMPUTING | 2020年 / 42卷 / 06期
关键词
Anderson acceleration; nonsmooth convex optimization; parallel and distributed optimization; proximal oracles; stabilization; global convergence; pathological settings; ALTERNATING DIRECTION METHOD; CONVERGENCE; OPTIMIZATION; ALGORITHM; EQUATIONS;
D O I
10.1137/19M1290097
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
We consider the problem of nonsmooth convex optimization with linear equality constraints, where the objective function is only accessible through its proximal operator. This problem arises in many different fields such as statistical learning, computational imaging, telecommunications, and optimal control. To solve it, we propose an Anderson accelerated Douglas-Rachford splitting (A2DR) algorithm, which we show either globally converges or provides a certificate of infeasibility/unboundedness under very mild conditions. Applied to a block separable objective, A2DR partially decouples so that its steps may be carried out in parallel, yielding an algorithm that is fast and scalable to multiple processors. We describe an open-source implementation and demonstrate its performance on a wide range of examples.
引用
收藏
页码:A3560 / A3583
页数:24
相关论文
共 50 条
  • [41] ADAPTIVE DOUGLAS-RACHFORD SPLITTING ALGORITHM FROM A YOSIDA APPROXIMATION STANDPOINT
    Liu, Zihan
    Ramchandran, Kannan
    SIAM JOURNAL ON OPTIMIZATION, 2021, 31 (03) : 1971 - 1998
  • [42] Tight Linear Convergence Rate Bounds for Douglas-Rachford Splitting and ADMM
    Giselsson, Pontus
    2015 54TH IEEE CONFERENCE ON DECISION AND CONTROL (CDC), 2015, : 3305 - 3310
  • [43] ON WEAK CONVERGENCE OF THE DOUGLAS-RACHFORD METHOD
    Svaiter, B. F.
    SIAM JOURNAL ON CONTROL AND OPTIMIZATION, 2011, 49 (01) : 280 - 287
  • [44] A note on the Douglas-Rachford splitting method for optimization problems involving hypoconvex functions
    Guo, Ke
    Han, Deren
    JOURNAL OF GLOBAL OPTIMIZATION, 2018, 72 (03) : 431 - 441
  • [45] 广义循环Douglas-Rachford算法
    郭科
    张有才
    西华师范大学学报(自然科学版), 2018, 39 (04) : 404 - 409
  • [46] On the local convergence of the Douglas-Rachford algorithm
    Bauschke, H. H.
    Noll, D.
    ARCHIV DER MATHEMATIK, 2014, 102 (06) : 589 - 600
  • [47] The cyclic Douglas-Rachford algorithm with r-sets-Douglas-Rachford operators
    Aragon Artacho, Francisco J.
    Censor, Yair
    Gibali, Aviv
    OPTIMIZATION METHODS & SOFTWARE, 2019, 34 (04): : 875 - 889
  • [48] A Control-Theoretic Approach to Analysis and Parameter Selection of Douglas-Rachford Splitting
    Seidman, Jacob H.
    Fazlyab, Mahyar
    Preciado, Victor M.
    Pappas, George J.
    IEEE CONTROL SYSTEMS LETTERS, 2020, 4 (01): : 199 - 204
  • [49] The Douglas-Rachford algorithm for a hyperplane and a doubleton
    Bauschke, Heinz H.
    Dao, Minh N.
    Lindstrom, Scott B.
    JOURNAL OF GLOBAL OPTIMIZATION, 2019, 74 (01) : 79 - 93
  • [50] Distributed Solution of GNEP over Networks via the Douglas-Rachford Splitting Method
    Huang, Yuanhanqing
    Hu, Jianghai
    2021 60TH IEEE CONFERENCE ON DECISION AND CONTROL (CDC), 2021, : 3110 - 3116