Grassmann algorithms for low rank approximation of matrices with missing values

DOI10.1007/s10543-010-0253-9zbMath1186.65050OpenAlexW2045123130MaRDI QIDQ2379362

Lars Eldén, Lennart Simonsson

Publication date: 19 March 2010

Published in: BIT (Search for Journal in Brave)

Full work available at URL: http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-54404

zbMATH Keywords

numerical examplessingular value decompositionGrassmann manifoldleast squaresJacobi-Davidson methodNewton's algorithmmissing elementslow rank approximation of matrices

Mathematics Subject Classification ID

Numerical solutions to overdetermined systems, pseudoinverses (65F20) Matrix completion problems (15A83)

Related Items

Two Newton methods on the manifold of fixed-rank matrices endowed with Riemannian quotient geometries ⋮ Fixed-rank matrix factorizations and Riemannian low-rank optimization ⋮ Robust principal component pursuit via inexact alternating minimization on matrix manifolds

Uses Software

• JDQR
• NewtonLib
• JDQZ
• ElemStatLearn

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Optimization and dynamical systems
• A Jacobi--Davidson Type SVD Method
• Adaptive Eigenvalue Computations Using Newton's Method on the Grassmann Manifold
• A Geometric Newton Method for Oja's Vector Field
• The Geometry of Algorithms with Orthogonality Constraints
• Templates for the Solution of Algebraic Eigenvalue Problems
• Optimization algorithms exploiting unitary constraints
• The geometry of weighted low-rank approximations
• The elements of statistical learning. Data mining, inference, and prediction