A parameterized Douglas–Rachford splitting algorithm for nonconvex optimization
作者：
Highlights：
• The parameterized Douglas–Rachford splitting algorithm provides great flexibility for practical applications. We established the global convergence of PDR splitting method in a nonconvex setting. Moreover, we provide a convergence analysis for an adaptive DR algorithm in nonconvex settings as a byproduct.
• Numerical experiments on nonconvex optimization problems such as sparse signal recovery nonconvex set feasibility problems and low rank matrix completion show the superiority compared to the other methods.
摘要
•The parameterized Douglas–Rachford splitting algorithm provides great flexibility for practical applications. We established the global convergence of PDR splitting method in a nonconvex setting. Moreover, we provide a convergence analysis for an adaptive DR algorithm in nonconvex settings as a byproduct.•Numerical experiments on nonconvex optimization problems such as sparse signal recovery nonconvex set feasibility problems and low rank matrix completion show the superiority compared to the other methods.
论文关键词：Parameterized Douglas–Rachford splitting method,Nonconvex optimization problems,Global convergence,Sparsity constrained least squares problem,Low rank matrix completion,Feasibility problem
论文评审过程：Received 28 August 2020, Revised 4 February 2021, Accepted 31 May 2021, Available online 23 June 2021, Version of Record 23 June 2021.
论文官网地址：https://doi.org/10.1016/j.amc.2021.126425
