Improved convergence rates and trajectory convergence for primal-dual dynamical systems with vanishing damping

被引:26
作者
Bot, Radu Ioan [1 ]
Nguyen, Dang-Khoa [1 ]
机构
[1] Univ Vienna, Fac Math, Oskar Morgenstern Pl 1, A-1090 Vienna, Austria
基金
奥地利科学基金会;
关键词
Augmented Lagrangian method; Primal-dual dynamical system; Damped inertial dynamics; Nesterov's accelerated gradient method; Convergence rates; Trajectory convergence; INERTIAL GRADIENT DYNAMICS; DISTRIBUTED ALGORITHMS; OPTIMIZATION; EQUATION;
D O I
10.1016/j.jde.2021.09.021
中图分类号
O1 [数学];
学科分类号
0701 ; 070101 ;
摘要
In this work, we approach the minimization of a continuously differentiable convex function under linear equality constraints by a second-order dynamical system with asymptotically vanishing damping term. The system is formulated in terms of the augmented Lagrangian associated to the minimization problem. We show fast convergence of the primal-dual gap, the feasibility measure, and the objective function value along the generated trajectories. In case the objective function has Lipschitz continuous gradient, we show that the primal-dual trajectory asymptotically weakly converges to a primal-dual optimal solution of the underlying minimization problem. To the best of our knowledge, this is the first result which guarantees the convergence of the trajectory generated by a primal-dual dynamical system with asymptotic vanishing damping. Moreover, we will rediscover in case of the unconstrained minimization of a convex differentiable function with Lipschitz continuous gradient all convergence statements obtained in the literature for Nesterov's accelerated gradient method. (c) 2021 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
引用
收藏
页码:369 / 406
页数:38
相关论文
共 57 条
[1]   Newton-Like Dynamics and Forward-Backward Methods for Structured Monotone Inclusions in Hilbert Spaces [J].
Abbas, B. ;
Attouch, H. ;
Svaiter, Benar F. .
JOURNAL OF OPTIMIZATION THEORY AND APPLICATIONS, 2014, 161 (02) :331-360
[3]   A second-order gradient-like dissipative dynamical system with Hessian-driven damping. Application to optimization and mechanics [J].
Alvarez, F ;
Attouch, H ;
Bolte, J ;
Redont, P .
JOURNAL DE MATHEMATIQUES PURES ET APPLIQUEES, 2002, 81 (08) :747-779
[4]   Convergence rate of inertial Forward-Backward algorithm beyond Nesterov's rule [J].
Apidopoulos, Vassilis ;
Aujol, Jean-Francois ;
Dossal, Charles .
MATHEMATICAL PROGRAMMING, 2020, 180 (1-2) :137-156
[5]   The heavy ball with friction method, I. The continuous dynamical system: Global exploration of the local minima of a real-valued function by asymptotic analysis of a dissipative dynamical system [J].
Attouch, H ;
Goudou, X ;
Redont, P .
COMMUNICATIONS IN CONTEMPORARY MATHEMATICS, 2000, 2 (01) :1-34
[6]  
Attouch H, 2009, PAC J OPTIM, V5, P17
[7]  
Attouch H., PURE APPL FUNC ANAL
[8]   FAST CONVEX OPTIMIZATION VIA INERTIAL DYNAMICS COMBINING VISCOUS AND HESSIAN-DRIVEN DAMPING WITH TIME RESCALING [J].
Attouch, Hedy ;
Balhag, Aicha ;
Chbani, Zaki ;
Riahi, Hassan .
EVOLUTION EQUATIONS AND CONTROL THEORY, 2022, 11 (02) :487-514
[9]   Fast Convergence of Dynamical ADMM via Time Scaling of Damped Inertial Dynamics [J].
Attouch, Hedy ;
Chbani, Zaki ;
Fadili, Jalal ;
Riahi, Hassan .
JOURNAL OF OPTIMIZATION THEORY AND APPLICATIONS, 2022, 193 (1-3) :704-736
[10]  
Attouch H, 2021, MINIMAX THEORY APPL, V6, P1