Matches in SemOpenAlex for { <https://semopenalex.org/work/W2918553587> ?p ?o ?g. }
- W2918553587 abstract "Matrix completion has attracted much interest in the past decade in machine learning and computer vision. For low-rank promotion in matrix completion, the nuclear norm penalty is convenient due to its convexity but has a bias problem. Recently, various algorithms using nonconvex penalties have been proposed, among which the proximal gradient descent (PGD) algorithm is one of the most efficient and effective. For the nonconvex PGD algorithm, whether it converges to a local minimizer and its convergence rate are still unclear. This work provides a nontrivial analysis on the PGD algorithm in the nonconvex case. Besides the convergence to a stationary point for a generalized nonconvex penalty, we provide more deep analysis on a popular and important class of nonconvex penalties which have discontinuous thresholding functions. For such penalties, we establish the finite rank convergence, convergence to restricted strictly local minimizer and eventually linear convergence rate of the PGD algorithm. Meanwhile, convergence to a local minimizer has been proved for the hard-thresholding penalty. Our result is the first shows that, nonconvex regularized matrix completion only has restricted strictly local minimizers, and the PGD algorithm can converge to such minimizers with eventually linear rate under certain conditions. Illustration of the PGD algorithm via experiments has also been provided. Code is available at this https URL." @default.
- W2918553587 created "2019-03-11" @default.
- W2918553587 creator A5032986088 @default.
- W2918553587 creator A5034251794 @default.
- W2918553587 creator A5048965499 @default.
- W2918553587 creator A5087982918 @default.
- W2918553587 date "2019-03-02" @default.
- W2918553587 modified "2023-09-27" @default.
- W2918553587 title "Matrix Completion via Nonconvex Regularization: Convergence of the Proximal Gradient Algorithm." @default.
- W2918553587 cites W1523985187 @default.
- W2918553587 cites W1675061566 @default.
- W2918553587 cites W195665135 @default.
- W2918553587 cites W1967138577 @default.
- W2918553587 cites W1969387119 @default.
- W2918553587 cites W1969698720 @default.
- W2918553587 cites W1971398013 @default.
- W2918553587 cites W2011359124 @default.
- W2918553587 cites W2014237985 @default.
- W2918553587 cites W2025666718 @default.
- W2918553587 cites W2031906930 @default.
- W2918553587 cites W2032024719 @default.
- W2918553587 cites W2037964547 @default.
- W2918553587 cites W2047071281 @default.
- W2918553587 cites W2059283452 @default.
- W2918553587 cites W2065180801 @default.
- W2918553587 cites W2074682976 @default.
- W2918553587 cites W2109240917 @default.
- W2918553587 cites W2111297856 @default.
- W2918553587 cites W2116413942 @default.
- W2918553587 cites W2119367754 @default.
- W2918553587 cites W2120872934 @default.
- W2918553587 cites W2128336967 @default.
- W2918553587 cites W2129732816 @default.
- W2918553587 cites W2134332047 @default.
- W2918553587 cites W2137414835 @default.
- W2918553587 cites W2146130798 @default.
- W2918553587 cites W2148232185 @default.
- W2918553587 cites W2159514083 @default.
- W2918553587 cites W2189224591 @default.
- W2918553587 cites W2271288847 @default.
- W2918553587 cites W2339666411 @default.
- W2918553587 cites W2548019060 @default.
- W2918553587 cites W2566240941 @default.
- W2918553587 cites W2605217922 @default.
- W2918553587 cites W2611328865 @default.
- W2918553587 cites W2797638056 @default.
- W2918553587 cites W2886697509 @default.
- W2918553587 cites W2909877831 @default.
- W2918553587 cites W2963714933 @default.
- W2918553587 cites W2964156132 @default.
- W2918553587 cites W2972886717 @default.
- W2918553587 cites W3013880646 @default.
- W2918553587 cites W3098595037 @default.
- W2918553587 cites W3105393233 @default.
- W2918553587 cites W3147510171 @default.
- W2918553587 cites W317954863 @default.
- W2918553587 hasPublicationYear "2019" @default.
- W2918553587 type Work @default.
- W2918553587 sameAs 2918553587 @default.
- W2918553587 citedByCount "0" @default.
- W2918553587 crossrefType "posted-content" @default.
- W2918553587 hasAuthorship W2918553587A5032986088 @default.
- W2918553587 hasAuthorship W2918553587A5034251794 @default.
- W2918553587 hasAuthorship W2918553587A5048965499 @default.
- W2918553587 hasAuthorship W2918553587A5087982918 @default.
- W2918553587 hasConcept C106159729 @default.
- W2918553587 hasConcept C106487976 @default.
- W2918553587 hasConcept C11413529 @default.
- W2918553587 hasConcept C121332964 @default.
- W2918553587 hasConcept C126255220 @default.
- W2918553587 hasConcept C127162648 @default.
- W2918553587 hasConcept C134306372 @default.
- W2918553587 hasConcept C154945302 @default.
- W2918553587 hasConcept C158693339 @default.
- W2918553587 hasConcept C159985019 @default.
- W2918553587 hasConcept C162324750 @default.
- W2918553587 hasConcept C163716315 @default.
- W2918553587 hasConcept C17744445 @default.
- W2918553587 hasConcept C189237950 @default.
- W2918553587 hasConcept C191795146 @default.
- W2918553587 hasConcept C192562407 @default.
- W2918553587 hasConcept C199539241 @default.
- W2918553587 hasConcept C2776135515 @default.
- W2918553587 hasConcept C2777303404 @default.
- W2918553587 hasConcept C2778459887 @default.
- W2918553587 hasConcept C31258907 @default.
- W2918553587 hasConcept C33923547 @default.
- W2918553587 hasConcept C41008148 @default.
- W2918553587 hasConcept C50522688 @default.
- W2918553587 hasConcept C57869625 @default.
- W2918553587 hasConcept C62520636 @default.
- W2918553587 hasConcept C72134830 @default.
- W2918553587 hasConcept C92207270 @default.
- W2918553587 hasConceptScore W2918553587C106159729 @default.
- W2918553587 hasConceptScore W2918553587C106487976 @default.
- W2918553587 hasConceptScore W2918553587C11413529 @default.
- W2918553587 hasConceptScore W2918553587C121332964 @default.
- W2918553587 hasConceptScore W2918553587C126255220 @default.
- W2918553587 hasConceptScore W2918553587C127162648 @default.
- W2918553587 hasConceptScore W2918553587C134306372 @default.