Conjugate-gradient method for computing the Moore-Penrose inverse and rank of a matrix

A conjugate-gradient method is developed for computing the Moore-Penrose generalized inverseA of a matrix and the associated projectors, by using the least-square characteristics of both the method and the inverseA . Two dual algorithms are introduced for computing the least-square and the minimum-norm generalized inverses, as well asA . It is shown that (i) these algorithms converge for any starting approximation; (ii) if they are started from the zero matrix, they converge toA ; and (iii) the trace of a sequence of approximations multiplied byA is a monotone increasing function converging to the rank ofA. A practical way of compensating the self-correcting feature in the computation ofA is devised by using the duality of the algorithms. Comparison with Ben-Israel's method is made through numerical examples. The conjugate-gradient method has an advantage over Ben-Israel's method.After having completed the present paper, the author received from Professor M. R. Hestenes his paper entitledPseudo Inverses and Conjugate Gradients. This paper treated the same subject and appeared in Communications of the ACM, Vol. 18, pp. 40–43, 1975.     

The Moore-Penrose inverse for sums of matrices under rank additivity conditions

Some results on the Moore-Penrose inverse for sums of matrices under rank additivity conditions are revisited and some new consequences are presented. Their extensions to the weighted Moore-Penrose inverse of sums of matrices under rank additivity conditions are also considered.     

A numerical procedure for computing the Moore-Penrose inverse

An algorithm for computing the Moore-Penrose inverse of an arbitraryn×m real matrixA is presented which uses a Gram-Schmidt like procedure to form anA-orthogonal set of vectors which span the subspace perpendicular to the kernel ofA. This one procedure will work for any value ofn andm, and for any value of rank (A).     

EP matrices: computation of the Moore-Penrose inverse via factorizations

EP matrices are a wide class of matrices which, among other things, can be characterized through factorizations. In this paper we are using two factorization algorithms in order to compute and factorize the Moore-Penrose inverse of a singular EP matrix. For the implementation of the algorithms we make use of a Computer Algebra System such as Maple. The results given by the proposed algorithms are faster and more accurate than the built-in Maple function in both symbolic and numerical tests, using matrices of different dimensions.     

Iterative method for computing the Moore-Penrose inverse based on Penrose equations

An iterative algorithm for estimating the Moore-Penrose generalized inverse is developed. The main motive for the construction of the algorithm is simultaneous usage of Penrose equations (2) and (4). Convergence properties of the introduced method as well as their first-order and second-order error terms are considered. Numerical experiment is also presented.     

A Family of higher-order convergent iterative methods for computing the Moore-Penrose inverse

In this paper, we extend the iterative method for computing the inner inverse of a matrix proposed in Li and Li [W.G. Li, Z. Li, A family of iterative methods for computing the approximate inverse of a square matrix and inner inverse of a non-square matrix, Applied Mathematics and Computation 215 (2010) 3433-3442] to compute the Moore-Penrose inverse of a matrix, and show that the generated sequence converges to the Moore-Penrose inverse of a matrix in a higher order. The performance of the method is tested on some randomly generated matrices.     

A finite algorithm for computing the weighted Moore-Penrose inverse A+MN

This paper presents a finite algorithm for computing the weighted Moore-Penrose inverse of a rectangular matrix. This algorithm may be viewed as an extension of Decell's algorithm.     

Convexity of the inverse and Moore-Penrose inverse

Convexity properties of the inverse of positive definite matrices and the Moore-Penrose inverse of nonnegative definite matrices with respect to the partial ordering induced by nonnegative definiteness are studied. For the positive definite case null-space characterizations are derived, and lead naturally to a concept of strong convexity of a matrix function, extending the conventional concept of strict convexity. The positive definite results are shown to allow for a unified analysis of problems in reproducing kernel Hilbert space theory and inequalities involving matrix means. The main results comprise a detailed study of the convexity properties of the Moore-Penrose inverse, providing extensions and generalizations of all the earlier work in this area.     

The Moore-Penrose and group inverse of square matrices and the Jordan canonical form

This paper is a continuation of the paper [3], where the first two Penrose's equations are solved for a square matrix, which is transformed into the Jordan canonical form. Our main aim is to solve the other two Penrose's equations for the same class of matrices. We also find a block representation of the group inverse and the form of the corresponding Jordan matrix, by solving the corresponding set of equations.     

Numerical study on Moore-Penrose inverse of tensors via Einstein product

Numerical Algorithms - The notation of Moore-Penrose inverse of matrices has been extended from matrix space to even-order tensor space with Einstein product. In this paper, we give the numerical...     

Numerical algorithms for the Moore-Penrose inverse of a matrix: Direct methods

Summary Direct methods for computing the Moore-Penrose inverse of a matrix are surveyed, classified and tested. It is observed that the algorithms using matrix decompositions or bordered matrices are numerically more stable.     

An efficient method for computing the inverse of arrowhead matrices

In this paper we propose a simple and effective method to find the inverse of arrowhead matrices which often appear in wide areas of applied science and engineering such as wireless communications systems, molecular physics, oscillators vibrationally coupled with Fermi liquid, and eigenvalue problems. A modified Sherman–Morrison inverse matrix method is proposed for computing the inverse of an arrowhead matrix. The effectiveness of the proposed method is illustrated and numerical results are presented along with comparative results.     

Linear operators that preserve maximal column rank of boolean matrices

This paper concerns two notions of column rank of matrices over semirings; column rank and maximal column rank. These two notions are the same over fields but differ for matrices over certain semirings. We determine how much the maximal column rank is different from the column ran for all m×n matrices over many semirings. We also characterize the linear operators which preserve the maximal column rank of Boolean matrices.     

Line digraphs and the Moore-Penrose inverse

Various characterizations of line digraphs and of Boolean matrices possessing a Moore-Penrose inverse are used to show that a square Boolean matrix has a Moore-Penrose inverse if and only if it is the adjacency matrix of a line digraph. A similar relationship between a nonsquare Boolean matrix and a bipartite graph is also given.     

On the covariance of the Moore-Penrose inverse

Given a square complex matrix A with Moore-Penrose inverse A², we describe the class of invertible matrices T such that (TAT^-1)²=TA²T^-1.     

A p-adic algorithm for computing the inverse of integer matrices

A method for computing the inverse of an (n×n)$(n\times n)$ integer matrix A$A$ using p$p$-adic approximation is given. The method is similar to Dixon’s algorithm, but ours has a quadratic convergence rate. The complexity of this algorithm (without using FFT or fast matrix multiplication) is O(n⁴(logn)²)$O\left(n^4{(logn)}^2\right)$, the same as that of Dixon’s algorithm. However, experiments show that our method is faster. This is because our methods decrease the number of matrix multiplications but increase the digits of the components of the matrix, which suits modern CPUs with fast integer multiplication instructions.     

Reverse order law for the Moore-Penrose inverse

In this paper we present new results related to the reverse order law for the Moore-Penrose inverse of operators on Hilbert spaces. Some finite-dimensional results are extended to infinite-dimensional settings.     

A determinantal inequality involving the Moore-Penrose inverse

Let A be a rectangular matrix of complex numbers whose rows are partitioned into r arbitrary blocks:

The Moore-Penrose inverses of each of these blocks are used to form the matrix B = (A₁⁺,…, A_r⁺). It is shown that 0 ? det (AB) ? 1. This is a generalized version of Hadamard's inequality.     

二元Boolean矩阵的加权Moore-Penrose逆

主要研究了二元Boolean矩阵A的加权Moore-Penrose逆的存在性问题,给出了二元Boolean矩阵A的加权Moore-Penrose逆存在的一些充分必要条件,并讨论了加权Moore-Penrose逆存在时的若干等价刻画及惟一性问题.