Qiaohua Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Qiaohua Liu has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Yimin Wei
School of Mathematical Sciences and Shanghai Key Laboratory of Contemporary Applied Mathematics, Fudan University, Shanghai, P. R. of China
4 shared publications - Zhigang Jia
Jiangsu Normal University
4 shared publications Aijing Liu
Qufu Normal University
2 shared publicationsJiehui Gu
Department of Mathematics, Shanghai University and Newtouch Center for Mathematics, Shanghai, People’s Republic of China
2 shared publicationsLei Yao
Shanghai Univ, Dept Math, Shanghai 200444, Peoples R China
2 shared publicationsQian Zhang
Department of Mathematics, Shanghai University, Shanghai, People’s Republic of China
2 shared publications
Latest Publications
- 22 September 2026
- Published by Springer Nature in Calcolo
- Vol. 63 (4) , 32
Fast adaptive tubal-rank-revealing algorithm for t-product based tensor approximation
- 22 September 2026
- Published by Springer Nature in Calcolo
- Vol. 63 (4) , 32
Fast adaptive tubal-rank-revealing algorithm for t-product based tensor approximation
- 6 August 2026
- Published by Wiley in Numerical Linear Algebra with Applications
- Vol. 33 (4)
Improved Analysis of Quaternion Randomized Block Krylov Iterations for Low‐Rank Matrix Approximation
This paper focuses on approximating a quaternion matrix and its singular values via the quaternion randomized block Krylov iteration (qRBKI) algorithm. Specifically, for qRBKI with block size , we establish novel gap‐based error bounds for the rank‐ matrix approximation in both Frobenius and spectral norms, explicitly demonstrating that qRBKI outperforms the quaternion randomized subspace iteration (qRSI) in approximation quality. For , leveraging simulated starting block vectors and quaternion polynomials, we show that the approximation error depends on the minimal relative singular value gap under the ‐order neighbouring structure. The near‐zero‐gap‐induced dependence can be eliminated by perturbing the input matrix with a low‐intensity random quaternion Gaussian or uniform matrix. Finally, numerical experiments on synthetic data and color image applications validate our theoretical findings and illustrate the effectiveness of qRBKI.
- 6 August 2026
- Published by Wiley in Numerical Linear Algebra with Applications
- Vol. 33 (4)
Improved Analysis of Quaternion Randomized Block Krylov Iterations for Low‐Rank Matrix Approximation
This paper focuses on approximating a quaternion matrix and its singular values via the quaternion randomized block Krylov iteration (qRBKI) algorithm. Specifically, for qRBKI with block size , we establish novel gap‐based error bounds for the rank‐ matrix approximation in both Frobenius and spectral norms, explicitly demonstrating that qRBKI outperforms the quaternion randomized subspace iteration (qRSI) in approximation quality. For , leveraging simulated starting block vectors and quaternion polynomials, we show that the approximation error depends on the minimal relative singular value gap under the ‐order neighbouring structure. The near‐zero‐gap‐induced dependence can be eliminated by perturbing the input matrix with a low‐intensity random quaternion Gaussian or uniform matrix. Finally, numerical experiments on synthetic data and color image applications validate our theoretical findings and illustrate the effectiveness of qRBKI.
- 15 April 2026
- Published by Springer Nature in BIT Numerical Mathematics
- Vol. 66 (2) , 28
Leverage score-based quaternion CUR decomposition: Gap error analysis and applications
- 15 April 2026
- Published by Springer Nature in BIT Numerical Mathematics
- Vol. 66 (2) , 28
Leverage score-based quaternion CUR decomposition: Gap error analysis and applications
- 24 February 2026
- Published by Wiley in Numerical Linear Algebra with Applications
- Vol. 33 (1)
Randomized Quaternion Generalized Singular Value Decomposition Algorithms With Error Analysis
In high‐dimensional data processing, quaternion generalized singular value decomposition (QGSVD) has become one of the most important solvers for advanced mathematical models. However, there are currently fewer efficient algorithms for calculating partial quaternion generalized singular values and vectors. When dealing with a quaternion matrix pair of substantial size, we present a stable QGSVD and two randomized algorithms cooperating with structure‐preserving and quaternion sampling strategies, and generate good low‐rank approximations to the original quaternion matrix pair. Explicitly, by leveraging random sampling based on the quaternion normal distribution for the original quaternion matrix pair, we propose both fixed‐rank and adaptive randomized algorithms for QGSVD. We also establish different kinds of error bounds for the randomized QGSVD approximation in the theoretical analysis and illustrate the effectiveness of the randomized QGSVD algorithms by numerical experiments on simulation data and practical color face recognition.
- 24 February 2026
- Published by Wiley in Numerical Linear Algebra with Applications
- Vol. 33 (1)
Randomized Quaternion Generalized Singular Value Decomposition Algorithms With Error Analysis
In high‐dimensional data processing, quaternion generalized singular value decomposition (QGSVD) has become one of the most important solvers for advanced mathematical models. However, there are currently fewer efficient algorithms for calculating partial quaternion generalized singular values and vectors. When dealing with a quaternion matrix pair of substantial size, we present a stable QGSVD and two randomized algorithms cooperating with structure‐preserving and quaternion sampling strategies, and generate good low‐rank approximations to the original quaternion matrix pair. Explicitly, by leveraging random sampling based on the quaternion normal distribution for the original quaternion matrix pair, we propose both fixed‐rank and adaptive randomized algorithms for QGSVD. We also establish different kinds of error bounds for the randomized QGSVD approximation in the theoretical analysis and illustrate the effectiveness of the randomized QGSVD algorithms by numerical experiments on simulation data and practical color face recognition.
- Open Access
Efficient adaptive randomized algorithms for fixed-threshold low-rank matrix approximation
The low-rank matrix approximation problems within a threshold are widely applied in information retrieval, image processing, background estimation of the video sequence problems and so on. This paper presents an adaptive randomized rank-revealing algorithm of the data matrix $A$, in which the basis matrix $Q$ of the approximate range space is adaptively built block by block, through a recursive deflation procedure on $A$. Detailed analysis of randomized projection schemes are provided to analyze the numerical rank reduce during the deflation. The provable spectral and Frobenius error $(I-QQ^T)A$ of the approximate low-rank matrix $\tilde A=QQ^TA$ are presented, as well as the approximate singular values. This blocked deflation technique is pass-efficient and can accelerate practical computations of large matrices. Applied to image processing and background estimation problems, the blocked randomized algorithm behaves more reliable and more efficient than the known Lanczos-based method and a rank-revealing algorithm proposed by Lee, Li and Zeng (in SIAM J. Matrix Anal. Appl. 31 (2009), pp. 503-525).Open AccessEfficient adaptive randomized algorithms for fixed-threshold low-rank matrix approximation
The low-rank matrix approximation problems within a threshold are widely applied in information retrieval, image processing, background estimation of the video sequence problems and so on. This paper presents an adaptive randomized rank-revealing algorithm of the data matrix $A$, in which the basis matrix $Q$ of the approximate range space is adaptively built block by block, through a recursive deflation procedure on $A$. Detailed analysis of randomized projection schemes are provided to analyze the numerical rank reduce during the deflation. The provable spectral and Frobenius error $(I-QQ^T)A$ of the approximate low-rank matrix $\tilde A=QQ^TA$ are presented, as well as the approximate singular values. This blocked deflation technique is pass-efficient and can accelerate practical computations of large matrices. Applied to image processing and background estimation problems, the blocked randomized algorithm behaves more reliable and more efficient than the known Lanczos-based method and a rank-revealing algorithm proposed by Lee, Li and Zeng (in SIAM J. Matrix Anal. Appl. 31 (2009), pp. 503-525). - 14 July 2025
- Published by Springer Nature in Indian Journal of Pure and Applied Mathematics
- Vol. 56 (3) , 989-1004
A Contribution to Condition Numbers of Quaternion Least Squares Problems
- 14 July 2025
- Published by Springer Nature in Indian Journal of Pure and Applied Mathematics
- Vol. 56 (3) , 989-1004
A Contribution to Condition Numbers of Quaternion Least Squares Problems
- 28 August 2024
- Published by Wiley in Numerical Linear Algebra with Applications
- Vol. 32 (1)
On rank‐revealing QR factorizations of quaternion matrices
This work develops theories and algorithms for computing rank‐revealing QR factorizations (qRRQR) of quaternion matrices. First, by introducing a novel quaternion determinant, a quasi‐Cramer's rule is established to investigate the existence theory of the qRRQR factorization of an quaternion matrix . The proposed theory provides a systematic approach for selecting linearly independent columns of in order to ensure that the resulting diagonal blocks , of the R‐factor possess rank revealing properties in both spectral and Frobenius norms. Secondly, by increasing the quaternion determinants of in each iteration by a factor of at least (), a greedy algorithm with cyclic block pivoting strategy is derived to implement the qRRQR factorization. This algorithm costs about real floating‐point operations, which is as nearly efficient as quaternion QR with column pivoting for most problems. It also shows the effectiveness and reliability, particularly when dealing with a class of quaternion Kahan matrices. Furthermore, two improved greedy algorithms are proposed. Experiments evaluations are conducted on synthetic data as well as color image compression and quaternion signals denoising. The experimental results validate the usefulness and efficiency of our improved algorithm across various scenarios.
- 28 August 2024
- Published by Wiley in Numerical Linear Algebra with Applications
- Vol. 32 (1)
On rank‐revealing QR factorizations of quaternion matrices
This work develops theories and algorithms for computing rank‐revealing QR factorizations (qRRQR) of quaternion matrices. First, by introducing a novel quaternion determinant, a quasi‐Cramer's rule is established to investigate the existence theory of the qRRQR factorization of an quaternion matrix . The proposed theory provides a systematic approach for selecting linearly independent columns of in order to ensure that the resulting diagonal blocks , of the R‐factor possess rank revealing properties in both spectral and Frobenius norms. Secondly, by increasing the quaternion determinants of in each iteration by a factor of at least (), a greedy algorithm with cyclic block pivoting strategy is derived to implement the qRRQR factorization. This algorithm costs about real floating‐point operations, which is as nearly efficient as quaternion QR with column pivoting for most problems. It also shows the effectiveness and reliability, particularly when dealing with a class of quaternion Kahan matrices. Furthermore, two improved greedy algorithms are proposed. Experiments evaluations are conducted on synthetic data as well as color image compression and quaternion signals denoising. The experimental results validate the usefulness and efficiency of our improved algorithm across various scenarios.
- Open Access
- 27 May 2024
- Published by Springer Nature in Calcolo
- Vol. 61 (2) , 1
Correction to: A Gauss–Newton method for mixed least squares-total least squares problems
- 27 May 2024
- Published by Springer Nature in Calcolo
- Vol. 61 (2) , 1
Open AccessCorrection to: A Gauss–Newton method for mixed least squares-total least squares problems
- 1 March 2024
- Published by Springer Nature in Calcolo
- Vol. 61 (1) , 1-27
A Gauss–Newton method for mixed least squares-total least squares problems
- 1 March 2024
- Published by Springer Nature in Calcolo
- Vol. 61 (1) , 1-27
A Gauss–Newton method for mixed least squares-total least squares problems
- Open Access
- 26 September 2023
- Published by University of Wyoming Libraries in The Electronic Journal of Linear Algebra
- Vol. 39, 516-532
On condition numbers of quaternion matrix inverse and quaternion linear systems with multiple right-hand sides
This paper is devoted to the condition numbers of quaternion linear system with multiple right-hand sides and the associated condition numbers of the quaternion matrix inverse as well. The explicit expressions of the unstructured and structured normwise, mixed, and componentwise condition numbers for the system are given. To reduce the computational cost of the condition numbers,compact and tight upper bounds for these condition numbers are proposed. For general sparse and badly scaled problems, numerical examples show that mixed and componentwise condition numbers are preferred than the normwise condition number for estimating the forward error of the solution, and structured condition numbers are tighter than the unstructured ones for some specific structured problems.
- 26 September 2023
- Published by University of Wyoming Libraries in The Electronic Journal of Linear Algebra
- Vol. 39, 516-532
Open AccessOn condition numbers of quaternion matrix inverse and quaternion linear systems with multiple right-hand sides
This paper is devoted to the condition numbers of quaternion linear system with multiple right-hand sides and the associated condition numbers of the quaternion matrix inverse as well. The explicit expressions of the unstructured and structured normwise, mixed, and componentwise condition numbers for the system are given. To reduce the computational cost of the condition numbers,compact and tight upper bounds for these condition numbers are proposed. For general sparse and badly scaled problems, numerical examples show that mixed and componentwise condition numbers are preferred than the normwise condition number for estimating the forward error of the solution, and structured condition numbers are tighter than the unstructured ones for some specific structured problems.