Xuefeng Duan

Dr. Xuefeng Duan


Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
20 December 2024 in Mathematics
Verified
Journal Article
Open access
PDF file
SDShan-Qi Duan
Shan-Qi Duan
Qing-Wen WangQing-Wen Wang
Qing-Wen Wang
XDXuefeng Duan
Xuefeng Duan
The application of eigenvalue theory to dual quaternion Hermitian matrices holds significance in the realm of multi-agent formation control. In this paper, we study the use of Rayleigh quotient iteration (RQI) for solving the right eigenpairs of dual quaternion Hermitian matrices. Combined with dual representation, the RQI algorithm can effectively compute the eigenvalue along with the associated eigenvector of the dual quaternion Hermitian matrices. Furthermore, by utilizing the minimal residual property of the Rayleigh quotient, a convergence analysis of the Rayleigh quotient iteration is derived. Numerical examples are provided to illustrate the high accuracy and low CPU time cost of the proposed Rayleigh quotient iteration compared with the power method for solving the dual quaternion Hermitian eigenvalue problem.

15 April 2013 in Linear and Multilinear Algebra
Verified
Journal Article
Closed access
XDXuefeng Duan
Xuefeng Duan
Qing-Wen WangQing-Wen Wang
Qing-Wen Wang
CLChun-Mei Li
Chun-Mei Li
In this paper, we consider a new nonlinear matrix equation , which is a special stochastic algebraic Riccati equation arising in stochastic control theory. Based on Brower’s fixed point theorem and Bhaskar–Lakshmikantham’s fixed point theroem, we derive some new sufficient conditions for the existence of a (unique) positive definite solution, which are easy to check. An iterative method is proposed to compute the unique positive definite solution. The error estimation of the iterative method is also given. Numerical examples show that the iterative method is feasible.

1 September 2013 in Linear and Multilinear Algebra
Verified
Journal Article
Closed access
XDXuefeng Duan
Xuefeng Duan
Qing-Wen WangQing-Wen Wang
Qing-Wen Wang
ALAnping Liao
Anping Liao
In this article, we consider the nonlinear matrix equation arising in an interpolation problem. We use the Thompson metric to prove that the matrix equation always has a unique positive definite solution. An iterative method is constructed to compute the unique positive definite solution and its error estimation formula is given. Based on the matrix differentiation, we give a precise perturbation bound for the unique positive definite solution. The new results are illustrated by some numerical examples in the end.

1 February 2019 in The Electronic Journal of Linear Algebra
Verified
Journal Article
Open access
PDF file
LCLei Cao
Lei Cao
ZCZhi Chen
Zhi Chen
XDXuefeng Duan
Xuefeng Duan
SKSelcuk Koyuncu
Selcuk Koyuncu
HLHuilan Li
Huilan Li
Let $\Omega_n$ denote the convex polytope of all $n\times n$ doubly stochastic matrices, and $\omega_{n}$ denote the convex polytope of all $n\times n$ doubly substochastic matrices. For a matrix $A\in\omega_n$, define the sub-defect of $A$ to be the smallest integer $k$ such that there exists an $(n+k)\times(n+k)$ doubly stochastic matrix containing $A$ as a submatrix. Let $\omega_{n,k}$ denote the subset of $\omega_n$ which contains all doubly substochastic matrices with sub-defect $k$. For $\pi$ a permutation of symmetric group of degree $n$, the sequence of elements $a_{1\pi(1)},a_{2\pi(2)}, \ldots, a_{n\pi(n)}$ is called the diagonal of $A$ corresponding to $\pi$. Let $h(A)$ and $l(A)$ denote the maximum and minimum diagonal sums of $A\in \omega_{n,k}$, respectively. In this paper, existing results of $h$ and $l$ functions are extended from $\Omega_n$ to $\omega_{n,k}.$ In addition, an analogue of Sylvesters law of the $h$ function on $\omega_{n,k}$ is proved.

14 December 2022 in Applied Sciences
Verified
Journal Article
Open access
PDF file
GDGuoqang Deng
Guoqang Deng
MTMin Tang
Min Tang
YZYuhao Zhang
Yuhao Zhang
YHYing Huang
Ying Huang
XDXuefeng Duan
Xuefeng Duan
Artificial neural network (ANN) is powerful in the artificial intelligence field and has been successfully applied to interpret complex image data in the real world. Since the majority of images are commonly known as private with the information intended to be used by the owner, such as handwritten characters and face, the private constraints form a major obstacle in developing high-precision image classifiers which require access to a large amount of image data belonging to multiple users. State-of-the-art privacy-preserving ANN schemes often use full homomorphic encryption which result in a substantial overhead of computation and data traffic for the data owners, and are restricted to approximation models by low-degree polynomials which lead to a large accuracy loss of the trained model compared to the original ANN model in the plain domain. Consequently, it is still a huge challenge to train an ANN model in the encrypted-domain. To mitigate this problem, we propose a privacy-preserving ANN system for secure constructing image classifiers, named IPPNN, where the server is able to train an ANN-based classifier on the combined image data of all data owners without being able to observe any images using primitives, such as randomization and functional encryption. Our system achieves faster training time and supports lossless training. Moreover, IPPNN removes the need for multiple communications among data owners and servers. We analyze the security of the protocol and perform experiments on a large scale image recognition task. The results show that the IPPNN is feasible to use in practice while achieving high accuracy.

5 November 2020 in Quantum Information Processing
Verified
Journal Article
Closed access
PDF file
XDXuefeng Duan
Xuefeng Duan
CLChi-Kwong Li
Chi-Kwong Li
DPDiane Christine Pelejo
Diane Christine Pelejo