Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4638416 | Journal of Computational and Applied Mathematics | 2015 | 15 Pages |
Abstract
In this paper, we discuss the following two minimum rank matrix approximation problems in the spectral norm: (i)For given A=AH∈Cm×m,B∈Cm×n, determining X∈S1X∈S1, such that rank(X)=minY∈S1rank(Y),S1={Y=YH∈Cn×n:‖A−BYBH‖2=min}S1={Y=YH∈Cn×n:‖A−BYBH‖2=min}.(ii)For given A=−AH∈Cm×m,B∈Cm×n, determining X∈S2X∈S2, such that rank(X)=minY∈S2rank(Y),S2={Y=−YH∈Cn×n:‖A−BYBH‖2=min}S2={Y=−YH∈Cn×n:‖A−BYBH‖2=min}.By applying the norm-preserving dilation theorem, the Hermitian-type (skew-Hermitian-type) generalized singular value decomposition (HGSVD, SHGSVD), we characterize the expressions of the minimum rank and derive a general form of minimum rank (skew) Hermitian solutions to the matrix approximation problem.
Keywords
Related Topics
Physical Sciences and Engineering
Mathematics
Applied Mathematics
Authors
Dongmei Shen, Musheng Wei, Yonghui Liu,