Semidefinite Programming Based Preconditioning for More Robust Near-Separable Nonnegative Matrix Factorization

DOI10.1137/130940670zbMath1316.15015arXiv1310.2273OpenAlexW2001096515MaRDI QIDQ5252582

Nicolas Gillis, Stephen A. Vavasis

Publication date: 2 June 2015

Published in: SIAM Journal on Optimization (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1310.2273

zbMATH Keywords

algorithmsemidefinite programmingpreconditioningseparabilitylarge scaledocument classificationnonnegative matrix factorizationrobustness to noisesuccessive projection algorithmhyperspectral unmixingactive-set methodminimum volume ellipsoid

Mathematics Subject Classification ID

Factorization of matrices (15A23) Semidefinite programming (90C22) Preconditioners for iterative methods (65F08)

Related Items

Block Bregman Majorization Minimization with Extrapolation, Heuristics for exact nonnegative matrix factorization, Efficient preconditioning for noisy separable nonnegative matrix factorization problems by successive projection based low-rank approximations, Enhancing Pure-Pixel Identification Performance via Preconditioning, Refinement of Hottopixx Method for Nonnegative Matrix Factorization Under Noisy Separability, Coseparable Nonnegative Matrix Factorization, Unnamed Item, Separable Quaternion Matrix Factorization for Polarization Images, Entrywise limit theorems for eigenvectors of signal-plus-noise matrix models with weak signals, Assigning topics to documents by successive projections, Nonnegative Matrix Factorization Via Archetypal Analysis, Dimensionality-Dependent Generalization Bounds for k-Dimensional Coding Schemes, Maximum Volume Inscribed Ellipsoid: A New Simplex-Structured Matrix Factorization Framework via Facet Enumeration and Convex Optimization, Multi-block Bregman proximal alternating linearized minimization and its application to orthogonal nonnegative matrix factorization, Simplex-Structured Matrix Factorization: Sparsity-Based Identifiability and Provably Correct Algorithms, Semidefinite Programming Based Preconditioning for More Robust Near-Separable Nonnegative Matrix Factorization, Provably Robust Blind Source Separation of Linear-Quadratic Near-Separable Mixtures, Detecting Overlapping Communities in Networks Using Spectral Methods, Robustness analysis of preconditioned successive projection algorithm for general form of separable NMF problem

Uses Software

• CVX

Cites Work

• Unnamed Item
• Unnamed Item
• Minimum-volume enclosing ellipsoids and core sets
• On the complexity of approximating the maximal inscribed ellipsoid for a polytope
• Latent semantic indexing: A probabilistic analysis
• On Khachiyan's algorithm for the computation of minimum-volume enclosing ellipsoids
• Nonnegative matrix factorization for spectral data analysis
• Robustness Analysis of Hottopixx, a Linear Programming Model for Factoring Nonnegative Matrices
• Robust Near-Separable Nonnegative Matrix Factorization Using Linear Optimization
• Graph Implementations for Nonsmooth Convex Programs
• Extensions of Lipschitz mappings into a Hilbert space
• On the Complexity of Nonnegative Matrix Factorization
• Generalized equations and their solutions, part II: Applications to nonlinear programming
• A Convex Analysis Framework for Blind Separation of Non-Negative Sources
• Rounding of Polytopes in the Real Number Model of Computation
• Semidefinite Programming Based Preconditioning for More Robust Near-Separable Nonnegative Matrix Factorization
• Computation of Minimum-Volume Covering Ellipsoids
• Learning the parts of objects by non-negative matrix factorization
• Sparse and Unique Nonnegative Matrix Factorization Through Data Preprocessing
• Faster Subset Selection for Matrices and Applications
• Computing a nonnegative matrix factorization -- provably
• Linear convergence of a modified Frank–Wolfe algorithm for computing minimum-volume enclosing ellipsoids