Generalization of core-EP inverse for rectangular matrices (Q2661265)

\textit{O. M. Baksalary} and \textit{G. Trenkler} [Linear Multilinear Algebra 58, No. 5--6, 681--697 (2010; Zbl 1202.15009)] have studied a core inverse of a matrix. The core-EP inverse of a matrix \(A \in \mathbb{C}^{n \times n}\), when \(k = \mathrm{ind}(A)\), is the unique matrix \(X \in \mathbb{C}^{n \times n}\) (denoted by \(A^{\dag}\) ) defined by \[ XAX = X,~~ \mathrm{Ran}(X) = \mathrm{Ran}(X^*) = \mathrm{Ran}(A^k).\] The core-EP inverse of \(A\) can be represented as \(A^{\dag}= A^{D}A^{k} (A^{k})^{ \dag}\). It is known that the core-EP inverse for matrices of index one becomes the core inverse. The *core-EP inverse of \(A \in \mathbb{C}^{n \times n}\), when \(k = \mathrm{ind}(A)\), introduced as the dual core-EP inverse in [\textit{K. M. Prasad} and \textit{K. S. Mohana}, Linear Multilinear Algebra 62, No. 6, 792--802 (2014; Zbl 1306.15006)], is the unique matrix \(X \in \mathbb{C}^{n \times n}\) such that \[ XAX = X,~~ \mathrm{Ran}(X) = \mathrm{Ran}(X^*) = \mathrm{Ran}((A^k)^*).\] In this paper, the notion of the generalized core-EP (g-core-EP) inverse is defined as a generalization of well-known core-EP inverse to rectangular matrices. The extension is defined using the outer inverse and the Moore-Penrose inverse. Various basic properties, representations and characterizations of the g-core-EP inverse are investigated. The *g-core-EP inverse is introduced and investigated as the dual of g-core-EP inverse. Several integral and limit representations of the g-core-EP and *g-core-EP inverses are studied. Also, applications of the g-core-EP and *g-core-EP inverses in solving some restricted systems of linear equations are considered.