000183060 001__ 183060
000183060 005__ 20190316235542.0
000183060 020__ $$a978-1-4673-0183-1
000183060 0247_ $$2doi$$a10.1109/SSP.2012.6319655
000183060 02470 $$2ISI$$a000309943200047
000183060 037__ $$aCONF
000183060 245__ $$aMatrix ALPS: Accelerated Low Rank and Sparse Matrix Reconstruction
000183060 269__ $$a2012
000183060 260__ $$bIEEE$$c2012$$aNew York
000183060 300__ $$a4
000183060 336__ $$aConference Papers
000183060 520__ $$aWe propose Matrix ALPS for recovering a sparse plus low-rank decomposition of a matrix given its corrupted and incomplete linear measurements. Our approach is a first-order projected gradient method over non-convex sets, and it exploits a well-known memory-based acceleration technique. We theoretically characterize the convergence properties of Matrix ALPS using the stable embedding properties of the linear measurement operator. We then numerically illustrate that our algorithm outperforms the existing convex as well as non-convex state-of-the-art algorithms in computational efficiency without sacrificing stability.
000183060 700__ $$0245321$$g199236$$aKyrillidis, Anastasios
000183060 700__ $$aCevher, Volkan$$g199128$$0243957
000183060 7112_ $$dAugust, 2012$$cAnn Arbor, Michigan, USA$$aIEEE Statistical Signal Processing Workshop (SSP)
000183060 773__ $$t2012 Ieee Statistical Signal Processing Workshop (Ssp)$$q185-188
000183060 8564_ $$uhttps://infoscience.epfl.ch/record/183060/files/SSP_final.pdf$$zPublisher's version$$s1263459$$yPublisher's version
000183060 909C0 $$xU12179$$0252306$$pLIONS
000183060 909CO $$qGLOBAL_SET$$pconf$$ooai:infoscience.tind.io:183060$$pSTI
000183060 917Z8 $$x199236
000183060 917Z8 $$x231598
000183060 937__ $$aEPFL-CONF-183060
000183060 973__ $$rREVIEWED$$sPUBLISHED$$aEPFL
000183060 980__ $$aCONF