首页
>
期刊首页
>
数学研究通讯
>
2026年2期
>
Transfer of Improvement Strategies Between DRS and ADMM:A Unified Classification Framework
Transfer of Improvement Strategies Between DRS and ADMM:A Unified Classification Framework
摘要
Douglas-Rachford splitting(DRS)and the alternating direction me-thod of multipliers(ADMM)are two fundamental first-order methods for struc-tured convex optimization.Although derived from different viewpoints,ADMM can be interpreted as the application of DRS to the dual problem.Based on this structural equivalence,this paper studies how algorithmic improve-ment strategies can be transferred between the two methods.We classify trans-ferable strategies into three categories:exact operator-level transfer,parameter-driven transfer,and heuristic transfer.Representative examples including re-laxation,metric scaling,adaptive parameter updates,and residual balancing are discussed to illustrate the different levels of transferability.This perspec-tive provides a systematic way to understand the relationship between DRS and ADMM and clarifies how algorithmic ideas developed for one method may inform the design of variants of the other,offering a unified framework that both explains existing variants and guides the design of new ones.
年，卷(期)：
2026, 42(2)
分类号：
65D18
65K10
机标分类号：
TP311.5
S
H319.9
在线出版日期：
2026年7月14日
