Zeroing Neural Network for Pseudoinversion of an Arbitrary Time-Varying Matrix Based on Singular Value Decomposition

Article Computer Science, Artificial Intelligence

Solving Complex-Valued Time-Varying Linear Matrix Equations via QR Decomposition With Applications to Robotic Motion Tracking and on Angle-of-Arrival Localization

Vasilios N. Katsikis et al.

Summary: The article investigates a complex-valued time-varying linear matrix equation problem and proposes a solution based on CVTVQR decomposition, which has shown effectiveness in numerical simulations and two applications.

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS (2022)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Zeroing Neural Network With Fuzzy Parameter for Computing Pseudoinverse of Arbitrary Matrix

Vasilios N. Katsikis et al.

Summary: The correlation between fuzzy logic systems (FLS) and zeroing neural networks (ZNN) design was investigated. It was shown that the gain parameter in ZNN design can be dynamically adjusted through a properly defined FLS. Dynamical systems for time-varying rank-deficient matrices were proposed and simulation experiments were performed to confirm the superiority of the FLS for dynamic adjustment of gain parameters.

IEEE TRANSACTIONS ON FUZZY SYSTEMS (2022)

添加到收藏夹

Article Mathematics

On the Calculation of the Moore-Penrose and Drazin Inverses: Application to Fractional Calculus

Khosro Sayevand et al.

Summary: This paper presents a highly efficient iterative method for obtaining Moore-Penrose and Drazin inverses, with potential applications in numerical linear algebra. The method is compared with existing approaches, demonstrating superior performance and accuracy, as well as providing criteria for smooth and non-smooth solutions in the fractional sense. Additionally, prototype examples such as fractional elliptic Poisson and sub-diffusion equations are considered, offering insights for extending the results to other scientific areas.

MATHEMATICS (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Continuous-Time Varying Complex QR Decomposition via Zeroing Neural Dynamics

Vasilios N. Katsikis et al.

Summary: In this paper, a continuous-time model is proposed for solving the time-varying problem of QR decomposition using the zeroing neural dynamics method, utilizing time derivative information from a known real or complex matrix. Theoretical analysis and numerical experiments demonstrate the effectiveness and convergence of the proposed method for solving the time-varying QR decomposition problem in both real and complex cases.

NEURAL PROCESSING LETTERS (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

A Penalty Strategy Combined Varying-Parameter Recurrent Neural Network for Solving Time-Varying Multi-Type Constrained Quadratic Programming Problems

Zhijun Zhang et al.

Summary: This paper proposes a penalty strategy combined with a varying-parameter recurrent neural network to solve time-varying quadratic programming problems, which can handle problems with equality constraints as well as inequality and bounded constraints. Numerical simulation experiments demonstrate the effectiveness and accuracy of this method.

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS (2021)

添加到收藏夹

Article Computer Science, Software Engineering