File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1287/moor.2014.0698
- Scopus: eid_2-s2.0-84944786665
- WOS: WOS:000367895700003
- Find via
Supplementary
- Citations:
- Appears in Collections:
Article: A strictly contractive Peaceman-Rachford splitting method with logarithmic-quadratic proximal regularization for convex programming
Title | A strictly contractive Peaceman-Rachford splitting method with logarithmic-quadratic proximal regularization for convex programming |
---|---|
Authors | |
Keywords | Convex programming Convergence rate Iteration complexity Logarithmic-quadratic proximal regularization Peaceman-Rachford splitting method |
Issue Date | 2015 |
Citation | Mathematics of Operations Research, 2015, v. 40, n. 4, p. 842-858 How to Cite? |
Abstract | Recently, a strictly contractive Peaceman-Rachford splitting method (PRSM) was proposed for a separable convex minimization model whose variables are subject to some linear constraints and two additional generic constraints. In general, the strictly contractive PRSM requires to solve two constrained minimization subproblems at each iteration. In this paper, we consider the case where the additional constraints on variables are positive orthants and apply the well-developed logarithmicquadratic proximal (LQP) regularization to regularize the subproblems of the strictly contractive PRSM. A new algorithm in combination with the strictly contractive PRSM and the LQP regularization is thus proposed. The new algorithm only needs to solve two unconstrained subproblems at each iteration. An inexact version allowing the unconstrained subproblems to be solved approximately subject to certain inexactness criterion is also studied. We prove the global convergence and establish a worst-case convergence rate measured by the iteration complexity for both the exact and inexact versions of the new algorithm. |
Persistent Identifier | http://hdl.handle.net/10722/251129 |
ISSN | 2023 Impact Factor: 1.4 2023 SCImago Journal Rankings: 1.215 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Li, Min | - |
dc.contributor.author | Yuan, Xiaoming | - |
dc.date.accessioned | 2018-02-01T01:54:41Z | - |
dc.date.available | 2018-02-01T01:54:41Z | - |
dc.date.issued | 2015 | - |
dc.identifier.citation | Mathematics of Operations Research, 2015, v. 40, n. 4, p. 842-858 | - |
dc.identifier.issn | 0364-765X | - |
dc.identifier.uri | http://hdl.handle.net/10722/251129 | - |
dc.description.abstract | Recently, a strictly contractive Peaceman-Rachford splitting method (PRSM) was proposed for a separable convex minimization model whose variables are subject to some linear constraints and two additional generic constraints. In general, the strictly contractive PRSM requires to solve two constrained minimization subproblems at each iteration. In this paper, we consider the case where the additional constraints on variables are positive orthants and apply the well-developed logarithmicquadratic proximal (LQP) regularization to regularize the subproblems of the strictly contractive PRSM. A new algorithm in combination with the strictly contractive PRSM and the LQP regularization is thus proposed. The new algorithm only needs to solve two unconstrained subproblems at each iteration. An inexact version allowing the unconstrained subproblems to be solved approximately subject to certain inexactness criterion is also studied. We prove the global convergence and establish a worst-case convergence rate measured by the iteration complexity for both the exact and inexact versions of the new algorithm. | - |
dc.language | eng | - |
dc.relation.ispartof | Mathematics of Operations Research | - |
dc.subject | Convex programming | - |
dc.subject | Convergence rate | - |
dc.subject | Iteration complexity | - |
dc.subject | Logarithmic-quadratic proximal regularization | - |
dc.subject | Peaceman-Rachford splitting method | - |
dc.title | A strictly contractive Peaceman-Rachford splitting method with logarithmic-quadratic proximal regularization for convex programming | - |
dc.type | Article | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1287/moor.2014.0698 | - |
dc.identifier.scopus | eid_2-s2.0-84944786665 | - |
dc.identifier.volume | 40 | - |
dc.identifier.issue | 4 | - |
dc.identifier.spage | 842 | - |
dc.identifier.epage | 858 | - |
dc.identifier.eissn | 1526-5471 | - |
dc.identifier.isi | WOS:000367895700003 | - |
dc.identifier.issnl | 0364-765X | - |