Nonnegative matrix factorization of a correlation matrix

DOI10.1016/J.LAA.2009.01.004MaRDI QIDQ1025855zbMATH OpenOpenAlexFDO

Authors J. J. I. M. van Kan, Peter Sonneveld, Xinzheng Huang, Cornelis W. Oosterlee

Publication date 23 June 2009

Published in Linear Algebra and its Applications (Search for Journal in Brave)

Full work available at URL http://resolver.tudelft.nl/uuid:958cb7dd-fe2e-463b-bda3-40622f4cfaaf

zbMATH Keywords

algorithmcorrelation matrixcredit portfolioFrobenius normKullback-Leibler divergenceloss modellinglow-rank approximationnonnegative matrix factorizationnumerical experimentssymmetric positive-semidefinite matrix

Mathematics Subject Classification ID

Factorization of matrices (15A23) Positive matrices and their generalizations; cones of matrices (15B48) Measures of association (correlation, canonical correlation, etc.) (62H20) Portfolio theory (91G10) Credit risk (91G40)

Review text

A correlation matrix is a symmetric positive-semidefinite \(n \times n\) matrix \(C\) which has a unit main diagonal. The basic question is essentially concerned with finding a factorization of the type \(C \approx AA^T\) such that \(AA^T\) should be a low-rank approximation of \(C\) together with an additional nonnegativity constraint. First, in section 2, an application from financial engineering in which the non-negative matrix factorization arises is discussed. In section 3, a counterexample for the strict equality \(C= AA^T\) is given. Then, in section 4, an algorithm -- based on a two step procedure -- is developed for obtaining a non-negative matrix \(A\) which, moreover, aims at minimizing the Frobenius norm of the matrix \(C-AA^T\) subject to \(A_{ij} \geq 0\) and \((AA^T)_{ii} = 1\). Finally, two alternative norms, the weighted Frobenius norm or the Kullback-Leibler divergence are considered and numerical experiments based on the Frobenius norm are also presented in section 5.

Recommendations

• Computing a nearest correlation matrix with factor structure
• Nonnegative matrix factorization
• Nonnegative rank factorization -- a heuristic approach via rank reduction
• On reduced rank nonnegative matrix factorization for symmetric nonnegative matrices
• scientific article; zbMATH DE number 7404610

Cites work

• Adaptive integration for multi-factor portfolio credit loss models
• Algorithms and applications for approximate nonnegative matrix factorization
• Efficient rank reduction of correlation matrices
• scientific article; zbMATH DE number 961607 (Why is no real title available?)
• Importance sampling for portfolio credit risk
• Learning the parts of objects by non-negative matrix factorization
• Nonnegative matrix factorization for spectral data analysis
• On reduced rank nonnegative matrix factorization for symmetric nonnegative matrices
• On the parameterization of the CreditRisk\(^+\) model for estimating credit portfolio risk
• Optimal low-rank approximation to a correlation matrix
• Rank reduction of correlation matrices by majorization
• The Gradient Projection Method for Nonlinear Programming. Part I. Linear Constraints

Cited in

(3)

• Alternative gradient algorithms for computing the nearest correlation matrix
• Block relaxation and majorization methods for the nearest correlation matrix with factor structure
• Non-negative matrix factorization with fixed row and column sums

This page was built for publication: Nonnegative matrix factorization of a correlation matrix

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1025855)