An extension of the m-weak group inverse (or m-WGI) on the set of rectangular matrices is provided to solve some systems of matrix equations. The extension is termed as the W-weighted m-WGI (or W-m-WGI). The W-m-WGI presents a new, wider class of generalized inverses which involves some already defined generalized inverses, such as the m-WGI, W-weighted weak group, and W-weighted Drazin inverse. Basic properties and diverse characterizations are proved for W-m-WGI. Several expressions for computing W-m-WGI are proposed in terms of known generalized inverses and projectors, as well as its limit and integral representations. The W-m-WGI class is utilized to solve some linear matrix equations and express their general solutions. Some new properties of the weighted generalized group inverse and recognized properties of the W-weighted Drazin inverse are obtained as corollaries. Numerical and symbolic test examples are presented to verify the obtained results.
Citation: D. Mosić, P. S. Stanimirović, L. A. Kazakovtsev. The m-weak group inverse for rectangular matrices[J]. Electronic Research Archive, 2024, 32(3): 1822-1843. doi: 10.3934/era.2024083
[1] | Daochang Zhang, Dijana Mosić, Liangyun Chen . On the Drazin inverse of anti-triangular block matrices. Electronic Research Archive, 2022, 30(7): 2428-2445. doi: 10.3934/era.2022124 |
[2] | Miao Luo, Yousong Chen, Dawei Gao, Lijun Wang . Inversion study of vehicle frontal collision and front bumper collision. Electronic Research Archive, 2023, 31(2): 776-792. doi: 10.3934/era.2023039 |
[3] | Yongge Tian . Characterizations of matrix equalities involving the sums and products of multiple matrices and their generalized inverse. Electronic Research Archive, 2023, 31(9): 5866-5893. doi: 10.3934/era.2023298 |
[4] | Yan Dong . Local Hölder continuity of nonnegative weak solutions of inverse variation-inequality problems of non-divergence type. Electronic Research Archive, 2024, 32(1): 473-485. doi: 10.3934/era.2024023 |
[5] | Yony Raúl Santaria Leuyacc . Elliptic equations in $ \mathbb{R}^2 $ involving supercritical exponential growth. Electronic Research Archive, 2024, 32(9): 5341-5356. doi: 10.3934/era.2024247 |
[6] | Guo-Niu Han . On the existence of permutations conditioned by certain rational functions. Electronic Research Archive, 2020, 28(1): 149-156. doi: 10.3934/era.2020009 |
[7] | Xiaoping Fang, Youjun Deng, Wing-Yan Tsui, Zaiyun Zhang . On simultaneous recovery of sources/obstacles and surrounding mediums by boundary measurements. Electronic Research Archive, 2020, 28(3): 1239-1255. doi: 10.3934/era.2020068 |
[8] | Yiming Zhang, Zhiwei Pan, Shuyou Zhang, Na Qiu . Probabilistic invertible neural network for inverse design space exploration and reasoning. Electronic Research Archive, 2023, 31(2): 860-881. doi: 10.3934/era.2023043 |
[9] | Xinlin Cao, Huaian Diao, Jinhong Li . Some recent progress on inverse scattering problems within general polyhedral geometry. Electronic Research Archive, 2021, 29(1): 1753-1782. doi: 10.3934/era.2020090 |
[10] | Yang Gao, Qingzhong Ji . On the inverse stability of $ z^n+c $. Electronic Research Archive, 2025, 33(3): 1414-1428. doi: 10.3934/era.2025066 |
An extension of the m-weak group inverse (or m-WGI) on the set of rectangular matrices is provided to solve some systems of matrix equations. The extension is termed as the W-weighted m-WGI (or W-m-WGI). The W-m-WGI presents a new, wider class of generalized inverses which involves some already defined generalized inverses, such as the m-WGI, W-weighted weak group, and W-weighted Drazin inverse. Basic properties and diverse characterizations are proved for W-m-WGI. Several expressions for computing W-m-WGI are proposed in terms of known generalized inverses and projectors, as well as its limit and integral representations. The W-m-WGI class is utilized to solve some linear matrix equations and express their general solutions. Some new properties of the weighted generalized group inverse and recognized properties of the W-weighted Drazin inverse are obtained as corollaries. Numerical and symbolic test examples are presented to verify the obtained results.
Throughout this work, Cp×n denotes the set involving p×n matrices with complex entries, and, for A∈Cp×n, rank(A) is its rank, A∗ is its conjugate-transpose matrix, N(A) is its null space, and R(A) is its range. The index ind(A) of A∈Cp×p is the smallest nonnegative integer k for which the equality rank(Ak)=rank(Ak+1) is satisfied. The symbol I denotes the identity matrix of adequate size. Standard notations PS and PS,T denote, respectively, the orthogonal projector onto a subspace S and a projector onto S along T when Cp is equal to the direct sum of the subspaces S and T.
Several definitions and properties of generalized inverses which are upgraded in this research are given. The Moore-Penrose inverse of A∈Cp×n is uniquely determined A†=X∈Cn×p as the solution to well-known Penrose equations [1]:
A=AXA,X=XAX,AX=(AX)∗,XA=(XA)∗. |
If X satisfies only equation XAX=X, it is an outer inverse of A. The outer inverse of A which is uniquely determined by the null space S and the range T is labeled with A(2)T,S=X∈Cn×p and satisfies
XAX=X,N(X)=S,R(X)=T, |
where s≤r=rank(A) is the dimension of the subspace T⊆Cn, and p−s is the dimension of the subspace S⊆Cp.
The following notation will be used:
Cp,n;k:={(A,W):A∈Cp×n, W∈Cn×p∖{0} and k=max{ind(AW),ind(WA)}}. |
The notion of the Drazin inverse was extended to rectangular matrices in [2]. For selected (A,W)∈Cp,n;k, the W-weighted Drazin inverse AD,W=X∈Cp×n of A is uniquely determined by the matrix equations
XWAWX=X,AWX=XWA,(AW)k+1XW=(AW)k. |
Especially, if p=n and W=I, AD,I:=AD reduces to the Drazin inverse of A. Further, for ind(A)=1, AD:=A# becomes the group inverse of A. Recall that [2]
AD,W=A[(WA)D]2=[(AW)D]2A. |
The notion of the core-EP inverse, proposed in [3] for a square matrix, was generalized to a rectangular matrix in [4]. If (A,W)∈Cp,n;k, the W-weighted core-EP inverse of A is the unique solution =X∈Cp×n to
WAWX=PR(WA)k,R((AW)k)=R(X). |
In a special case p=n and W=I, becomes the core-EP inverse
of A. According to original definitions in [5] and [6,7,8], it is important to note
![]() |
and
![]() |
Some useful characterizations and representations of the core-EP inverse are presented in [3,9,10,11,12,13,14]. In the case ind(A)=1, reduces to the core inverse
=A#AA† of A [15].
The weak group inverse (WGI) was presented for a square matrix in [16] as an extension of the group inverse. The WGI is extended in [17] to a rectangular matrix and in [18] to Hilbert space operators. For (A,W)∈Cp,n;k, the W-weighted WGI (W-WGI) of A is the unique solution Aⓦ,W=X∈Cp×n of the system [17,18]
![]() |
and it is expressed by [17,18]
![]() |
When p=n and W=I, Aⓦ,I:=Aⓦ reduces to the WGI of A
![]() |
Remark that, for 1=ind(A), Aⓦ=A#. Useful results about WGI were given in [17,18,19,20,21,22,23].
The concept of the m-weak group inverse (m-WGI) was introduced in [24] as an extension of the WGI. Exactly, if m∈N, the m-WGI of A∈Cn×n is the unique matrix Aⓦm=X∈Cn×n such that [25]
![]() |
(1.1) |
Recall that
![]() |
Clearly, Aⓦ1=Aⓦ, and particularly Aⓦ2=()3A2 becomes the generalized group (GG) inverse of A, established in [26]. It is interesting that, if ind(A)≤m, Aⓦm=AD. Various properties of m-WGI were presented in [24,25,27,28].
Recent research about m-WGI as well as the fact that m-WGI is an important extension of the WGI, GG, Drazin inverse, and group inverse motivated us to further investigate this topic. The current popular trend in the research of generalized inverses consists in defining new generalized inverses that are based on suitable combinations of existing generalized inverses as well as in their application in solving appropriate systems of linear equations. Considering the system (1.1) for defining m-WGI, our first aim is to solve a system of matrix equations which is an extension of the system (1.1) from the square matrix case to an arbitrary case. Since the m-WGI is restricted to square matrices, our main goal is to extend this notion to W-m-WGI inverses on rectangular matrices. To solve a certain system of matrix equations on rectangular complex matrices, we extend the notions of m-WGI, W-WGI, and the W-weighted Drazin inverse by introducing a wider class of generalized inverses, termed as the W-weighted m-WGI (W-m-WGI) for a rectangular matrix. Particularly, an extension of the GG inverse on rectangular matrices is obtained. It is important to mention that we recover significant results for the W-weighted Drazin inverse in a particular case. A class of systems of linear equations is found that can be efficiently solved applying W-m-WGI. This results is an extension of known results about the W-weighted Drazin solution and the Drazin solution of exact linear systems.
The global structure of the work is based on sections with the following content. Several characterizations for the W-m-WGI are proved in Section 2 without and with projectors. We develop important expressions for the W-m-WGI based on core–EP, Drazin, and Moore-Penorse inverses of proper matrices. As a consequence, we introduce the weighted version of GG inverse and give its properties. Limit and integral formulae for computing the W-m-WGI are part of Section 3. Section 4 investigates applications of the W-m-WGI in solving specific matrix equations. Numerical experiments are presented in Section 5.
We introduce the W-weighted m-WGI on rectangular matrices as a class of generalized inverses that includes notions of the m-WGI and the W-weighted WGI.
Theorem 2.1. If (A,W)∈Cp,n;k and m∈N, then X=(W)m+1(AW)m−1A is the unique solution to the matrix system
![]() |
(2.1) |
Proof. Using the identity AWW
=
, the subsequent transformations are obtained for X:=(
W)m+1(AW)m−1A:
![]() |
which further leads to
![]() |
Hence, X=(W)m+1(AW)m−1A is a solution to (2.1).
An arbitrary solution X to the system (2.1) satisfies
![]() |
which leads to the conclusion that X=(W)m+1(AW)m−1A is the unique solution to (2.1).
Definition 2.1. Under such conditions (A,W)∈Cp,n;k and m∈N, the W-weighted m-WGI (shortly W-m-WGI) inverse of A is defined by the expression
![]() |
Several special appearance forms of the W-m-WGI show its importance and are listed as follows:
− when p=n and W=I, the I-m-WGI coincides with the m-WGI Aⓦm=()m+1Am;
− if m=1, then (W)2A=Aⓦ,W, that is, the W-1-WGI reduces to the W-WGI;
− for m=2, the W-2-WGI is introduced as Aⓦ2,W=(W)3AWA and presents an extension of the GG inverse;
− in the case k≤m, it follows Aⓦm,W=AD,W (see Lemma 2.1).
Some computationally useful representations of the W-m-WGI are developed in subsequent statements.
Lemma 2.1. If (A,W)∈Cp,n;k, m∈N and l≥k, then
![]() |
Furthermore, for m≥k, it follows that Aⓦm,W=AD,W.
Proof. First, by induction on m, notice that gives
![]() |
Further, based on
![]() |
and
(WA)k[(WA)k]†=PR((WA)k)=PR((WA)l)=(WA)l[(WA)l]†, |
we obtain
![]() |
In the case m≥k, it follows that
Aⓦm,W=A[(WA)D]m+2(WA)k[(WA)k]†(WA)m=A[(WA)D]m+2(WA)m=A[(WA)D]2=AD,W. |
Remark 2.1. Note that Aⓦm,W=(WA)ⓦm implies the interesting identity WAⓦm,W=PR((WA)l)(WA)ⓦm for l≥k. This last identity is an extension of the classical property of the W-weighted Drazin inverse WAD,W=(WA)D. About the dual property AD,WW=(AW)D, if the equality
A is satisfied (which is not true in general; for details see [5]), we can verify that Aⓦm,W=
(AW)ⓦmA and so Aⓦm,WW=
(AW)ⓦmAW.
Representations for the W-2-WGI and W-weighted Drazin inverse are obtained as consequences of Lemma 2.1 when m=2 or m=l≥k, respectively.
Corollary 2.1. If (A,W)∈Cp,n;k and l≥k, then
![]() |
and
![]() |
Notice that Corollary 2.1 recovers the known expressions for the W-weighted Drazin inverse [29,30].
In Lemma 2.2, we show that the W-m-WGI Aⓦm,W is an outer inverse of WAW and find its range and null spaces.
Lemma 2.2. If (A,W)∈Cp,n;k and m∈N, the following representations are valid:
(i) Aⓦm,W=(WAW)(2)R((AW)k),N([(WA)k]∗(WA)m);
(ii) WAWAⓦm,W=PR((WA)k),N([(WA)k]∗(WA)m);
(iii) Aⓦm,WWAW=PR((AW)k),N([(WA)k]∗(WA)m+1W).
Proof. (ⅰ) Based on Lemma 2.1 it follows that Aⓦm,W=A(WA)k[(WA)k+m+2]†(WA)m, which yields R(Aⓦm,W)⊆R((AW)k) and
Aⓦm,WWAWAⓦm,W=A(WA)k[(WA)k+m+2]†(WA)k+m+2[(WA)k+m+2]†(WA)m=A(WA)k[(WA)k+m+2]†(WA)m=Aⓦm,W. |
Another application of Lemma 2.1 yields
(AW)k=[(AW)D]m+2(AW)k+m+2=A[(WA)D]m+2(WA)k+m+1W=A[(WA)D]m+2(WA)k[(WA)k]†(WA)k+m+1W=Aⓦm,W(WA)k+1W | (2.2) |
and so R((AW)k)⊆R(Aⓦm,W). Thus, R(Aⓦm,W)=R((AW)k). Also,
N([(WA)k]∗(WA)m)=N([(WA)k+m+2]∗(WA)m)=N([(WA)k+m+2]†(WA)m)=N(A(WA)k[(WA)k+m+2]†(WA)m)=N(Aⓦm,W). |
(ⅱ) By the part (ⅰ), WAWAⓦm,W is a projector, and
N(WAWAⓦm,W)=N(Aⓦm,W)=N([(WA)k]∗(WA)m). |
The equalities WAWAⓦm,W=(WA)k+2[(WA)k+m+2]†(WA)m and
(WA)k=(WA)k+2[(WA)D]2=(WA)k+2[(WA)k+2]†(WA)k+2[(WA)D]2=(WA)k+2PR([(WA)k+2]∗)[(WA)D]2=(WA)k+2PR([(WA)m+k+2]∗)[(WA)D]2=(WA)k+2[(WA)m+k+2]†(WA)m+k+2[(WA)D]2=WAWAⓦm,W(WA)k+2[(WA)D]2=WAWAⓦm,W(WA)k |
imply R(WAWAⓦm,W)=R((WA)k).
(ⅲ) It is clear, by (ⅰ), that R(Aⓦm,WWAW)=R(Aⓦm,W)=R((AW)k). The identity N(Aⓦm,WWAW)=N([(WA)k]∗(WA)m+1W) is verified in a similar manner as in (ⅰ).
Remark 2.2. For A∈Cm×n, B∈Cq×t, C∈Cs×p, M∈Cp×m, and N∈Cn×q, by [31,32], the (M,N)-weighted (B,C)-inverse of A is represented by A(2,M,N)(B,C)=(MAN)(2)R(B),N(C). By Lemma 2.2, Aⓦm,W=(WAW)(2)R((AW)k),N([(WA)k]∗(WA)m), and thus Aⓦm,W is the (W,W)-weighted ((AW)k),[(WA)k]∗(WA)m)-inverse of A. Since the (B,C)-inverse of A is given as A(2)(B,C)=A(2)R(B),N(C) [33], it follows that Aⓦm,W=(WAW)(2)R((AW)k),N([(WA)k]∗(WA)m) is the ((AW)k,[(WA)k]∗(WA)m)-inverse of WAW.
Lemma 2.2 and the Urquhart formula [1] give the next representations for Aⓦm,W.
Corollary 2.2. If (A,W)∈Cp,n;k and m∈N, the W-m-WGI of A is represented as
Aⓦm,W=(AW)k([(WA)k]∗(WA)m+k+1W)†[(WA)k]∗(WA)m. |
If m=2 or m=k in Lemma 2.2 and Corollary 2.2, we obtain the next properties related to the W-2-WGI and W-weighted Drazin inverse.
Corollary 2.3. If (A,W)∈Cp,n;k, the following statements hold:
(i) Aⓦ2,W=(WAW)(2)R((AW)k),N([(WA)k]∗(WA)2)=(AW)k([(WA)k]∗(WA)k+3W)†[(WA)k]∗(WA)2;
(ii) WAWAⓦ2,W=PR((WA)k),N([(WA)k]∗(WA)2);
(iii) Aⓦ2,WWAW=PR((AW)k),N([(WA)k]∗(WA)3W);
(iv) AD,W=(WAW)(2)R((AW)k),N((WA)k)=(AW)k([(WA)k]∗(WA)2k+1W)†[(WA)k]∗(WA)k;
(v) WAWAD,W=PR((WA)k),N((WA)k);
(vi) AD,WWAW=PR((AW)k),N((AW)k).
Some necessary and sufficient conditions for a rectangular matrix to be the W-m-WGI are considered.
Theorem 2.2. If (A,W)∈Cp,n;k, X∈Cp×n, and m∈N, the subsequent statements are equivalent:
(i) X=Aⓦm,W;
(ii) AWX=(W)m(AW)m−1A and R(X)=R((AW)k);
(iii) AWX=(W)m(AW)m−1A and R(X)⊆R((AW)k);
(iv) AWXWX=X, X(WA)k+1W=(AW)k and [(WA)k]∗(WA)m+1WX=[(WA)k]∗(WA)m;
(v) XWAWX=X, R(X)=R((AW)k) and [(WA)k]∗(WA)m+1WX=[(WA)k]∗(WA)m;
(vi) XWAWX=X, AWX=(W)m(AW)m−1A and XWA=(
W)m+1(AW)mA;
(vii) XWAWX=X, WAWX=W(W)m(AW)m−1A and XWAW=(
W)m+1(AW)m+1;
(viii) XWAWX=X, AWXWA=(W)m(AW)mA, AWX=(
W)m(AW)m−1A and XWA=(
W)m+1(AW)mA;
(ix) XWAWX=X, WAWXWAW=W(W)m(AW)m+1, WAWX=W(
W)m(AW)m−1A and XWAW=(
W)m+1(AW)m+1;
(x) X=WAWX and AWX=(
W)m(AW)m−1A (or WAWX=W(
W)m(AW)m−1A);
(xi) X=AWWX and
WX=(
W)m+1(AW)m−1A;
(xii) X=AD,WWAWX and AWX=(W)m(AW)m−1A (or WAWX=W(
W)m(AW)m−1A);
(xiii) X=(W)m+1(AW)m+1X and AWX=(
W)m(AW)m−1A (or WAWX=W(
W)m(AW)m−1A);
(xiv) X=XW(W)m(AW)m−1A and XWA=(
W)m+1(AW)mA (or XWAW=(
W)m+1(AW)m+1).
Proof. (ⅰ) ⇒ (ⅱ): It follows from Theorem 2.1 and Lemma 2.2.
(ⅱ) ⇒ (ⅲ): This implication is obvious.
(ⅲ) ⇒ (ⅰ): Because R(X)⊆R((AW)k), we have
X=(AW)kU=(AW)k[(AW)k]†(AW)kU=(AW)k[(AW)k]†X, |
for some U∈Cp×n. Notice that, by AWX=(W)m(AW)m−1A,
![]() |
An application of Theorem 2.1 leads to the conclusion X=Aⓦm,W.
(ⅰ) ⇒ (ⅳ): For X=Aⓦm,W, Theorem 2.1 implies AWXWX=X. The equality (2.2) gives X(WA)k+1W=(AW)k. Using Lemma 2.1, we get X=A(WA)k[(WA)k+m+2]†(WA)m, which implies
[(WA)k]∗(WA)m+1WX=[(WA)k]∗(WA)m+1WA(WA)k[(WA)k+m+2]†(WA)m=[(WA)k]∗(WA)k+m+2[(WA)k+m+2]†(WA)m=[(WA)k]∗(WA)k[(WA)k]†(WA)m=[(WA)k]∗(WA)m. |
(ⅳ) ⇒ (ⅴ): Notice, by X(WA)k+1W=(AW)k and
X=AWXWX=(AW)2(XW)2X=⋯=(AW)r(XW)rX, | (2.3) |
for arbitrary r∈N, that R(X)=R((AW)k). Hence, X=(AW)kU, for some U∈Cp×n, and
XWAWX=XWAW(AW)kU=(X(WA)k+1W)U=(AW)kU=X. |
(ⅴ) ⇒ (ⅰ): The assumptions R(X)=R((AW)k) and XWAWX=X give (AW)k=XV=XWAW(XV)=XW(AW)k+1, for some V∈Cn×p. Since X=(AW)kU, for some U∈Cp×n, we get
AWXWX=AWXW(AW)kU=AW(XW(AW)k+1)(AW)DU=AW(AW)k(AW)DU=(AW)kU=X. |
The assumption [(WA)k]∗(WA)m+1WX=[(WA)k]∗(WA)m yields
(WA)k[(WA)k]†(WA)m+1WX=([(WA)k]†)∗[(WA)k]∗(WA)m+1WX=([(WA)k]†)∗[(WA)k]∗(WA)m=(WA)k[(WA)k]†(WA)m. |
Because (2.3) holds, we obtain
![]() |
Theorem 2.1 gives X=Aⓦm,W.
(ⅰ) ⇒ (ⅵ) ⇒ (ⅶ): These implications are clear.
(ⅶ) ⇒ (ⅰ): Using XWAWX=X, WAWX=W(W)m(AW)m−1A and XWAW=(
W)m+1(AW)m+1, we get
![]() |
(ⅵ) ⇔ (ⅷ) and (ⅶ) ⇔ (ⅸ): These equivalences are evident.
(ⅰ) ⇒ (ⅹ): From X=(W)m+1(AW)m−1A, we observe
![]() |
(ⅹ) ⇒ (ⅰ): Applying X=WAWX and AWX=(
W)m(AW)m−1A, we obtain X=
W(AWX)=
W(
W)m(AW)m−1A=(
W)m+1(AW)m−1A.
(v) ⇔ (xiv): This equivalence follows as (ⅰ) ⇔ (ⅶ).
We also characterize the W-m-WGI in the following two ways.
Theorem 2.3. If (A,W)∈Cp,n;k and m∈N, then
(i) Aⓦm,W is the unique solution to
WAWX=PR((WA)k),N([(WA)k]∗(WA)m)andR(X)⊆R((AW)k); | (2.4) |
(ii) Aⓦm,W is the unique solution to
XWAW=PR((AW)k),N([(WA)k]∗(WA)m+1W)andR(X∗)⊆R([(WA)m]∗(WA)k). | (2.5) |
Proof. (ⅰ) By Lemma 2.2, X=Aⓦm,W is a solution to (2.4). If the system (2.4) has two solutions X and X1, notice
WAW(X−X1)=PR((WA)k),N([(WA)k]∗(WA)m)−PR((WA)k),N([(WA)k]∗(WA)m)=0 |
and R(X−X1)⊆R((AW)k). Therefore,
R(X−X1)⊆N(WAW)∩R((AW)k)⊆N((AW)k)∩R((AW)k)={0}, |
i.e., X=X1 is the unique solution of the system of Eqs (2.4).
(ⅱ) Lemmas 2.1 and 2.2 imply validity of (2.5) for X=Aⓦm,W=A(WA)k[(WA)k+m+2]†(WA)m.
The assumption that two solutions X and X1 satisfy (2.5) leads to the conclusion
R(X∗−X∗1)⊆R([(WA)m]∗(WA)k)∩N((WAW)∗)⊆R((WAWAⓦm,W)∗)∩N((WAWAⓦm,W)∗)={0}, |
that is, X=X1.
Corresponding characterizations of the W-2-WGI and W-weighted Drazin inverse are derived as particular cases m=2 and m=k of Theorem 2.3, respectively.
Corollary 2.4. The following statements are valid for (A,W)∈Cp,n;k:
(i) Aⓦ2,W is the unique solution to
WAWX=PR((WA)k),N([(WA)k]∗(WA)2)andR(X)⊆R((AW)k); |
(ii) Aⓦ2,W is the unique solution to
XWAW=PR((AW)k),N([(WA)k]∗(WA)3W)andR(X∗)⊆R([(WA)2]∗(WA)k); |
(iii) AD,W is the unique solution to
WAWX=PR((WA)k),N((WA)k)andR(X)⊆R((AW)k); |
(iv) AD,W is the unique solution to
XWAW=PR((AW)k),N((AW)k)andR(X∗)⊆R([(WA)k]∗). |
Some formulae for the W-m-WGI are given in this section.
We present a relation between a nonsingular bordered matrix and the W-m-WGI. Precisely, by Theorem 3.1, when the inverse of a proper bordered matrix is known, then the corresponding position of that inverse gives the W-m-WGI.
Theorem 3.1. Let (A,W)∈Cp,n;k and m∈N. Assume that full-column rank matrices G and H∗ fulfill
N([(WA)k]∗(WA)m)=R(G)andR((AW)k)=N(H). |
Then,
N=[WAWGH0] |
is nonsingular, and
N−1=[Aⓦm,W(I−Aⓦm,WWAW)H†G†(I−WAWAⓦm,W)−G†(WAW−WAWAⓦm,WWAW)H†]. | (3.1) |
Proof. Lemma 2.2 gives Aⓦm,W=(WAW)(2)R((AW)k),N([(WA)k]∗(WA)m). Since
R(I−WAWAⓦm,W)=N(WAWAⓦm,W)=N(Aⓦm,W)=N([(WA)k]∗(WA)m)=R(G)=R(GG†)=N(I−GG†), |
we have (I−GG†)(I−WAWAⓦm,W)=0, that is, GG†(I−WAWAⓦm,W)=(I−WAWAⓦm,W). From R(Aⓦm,W)=R((AW)k)=N(H), we get HAⓦm,W=0. Set Y for the right hand side of (3.1). Then,
NY=[WAWAⓦm,W+GG†(I−WAWAⓦm,W)WAW(I−Aⓦm,WWAW)H†−GG†(I−WAWAⓦm,W)WAWH†HAⓦmH(I−AⓦmWAW)H†]=[WAWAⓦm,W+I−WAWAⓦm,W(I−WAWAⓦm,W)WAWH†−(I−WAWAⓦm,W)WAWH†0HH†]=[I00I]=I. |
and similarly YN=I. So, N is nonsingular with N−1=Y.
We investigate limit and integral expressions for W-m-WGI motivated by limit and integral formulae of known generalized inverses [34,35,36,37].
Theorem 3.2. If (A,W)∈Cp,n;k, m∈N, and l≥k, then
Aⓦm,W=limλ→0A(WA)l[(WA)l+m+2]∗((WA)l+m+2[(WA)l+m+2]∗+λI)−1(WA)m=limλ→0A(WA)l([(WA)l+m+2]∗(WA)l+m+2+λI)−1[(WA)l+m+2]∗(WA)m. |
Proof. Lemma 2.1 gives Aⓦm,W=A(WA)l[(WA)l+m+2]†(WA)m. According to the limit representation for the Moore-Penrose inverse given in [36], we derive
[(WA)l+m+2]†=limλ→0[(WA)l+m+2]∗((WA)l+m+2[(WA)l+m+2]∗+λI)−1=limλ→0([(WA)l+m+2]∗(WA)l+m+2+λI)−1[(WA)l+m+2]∗, |
which implies the rest.
Since W-m-WGI belongs to outer inverses, the limit representation of the outer inverse proposed in [35] implies the limit representation of the W-m-WGI.
Theorem 3.3. Let (A,W)∈Cp,n;k and m∈N. Suppose that H1∈Cp×ss, R(H1)=R((AW)k), H2∈Cs×ns, and n−s is the dimension of the subspace N(H2)=N([(WA)k]∗(WA)m) in Cn. Then,
Aⓦm=limv→0H1(vI+H2WAWH1)−1H2=limu→0(uI+H1H2WAW)−1H1H2=limv→0H1H2(vI+WAWH1H2)−1. |
Proof. Since Aⓦm,W=(WAW)(2)R((AW)k),N([(WA)k]∗(WA)m), by [35, Theorem 7], it follows that
Aⓦm,W=limu→0H1(uI+H2WAWH1)−1H2. |
Some integral formulae are established for the W-m-WGI.
Theorem 3.4. If (A,W)∈Cp,n;k, m∈N, and l≥k, then
Aⓦm,W=∫∞0A(WA)l[(WA)l+m+2]∗exp(−(WA)l+m+2[(WA)l+m+2]∗v)(WA)mdv. |
Proof. According to [34],
[(WA)l+m+2]†=∫∞0[(WA)l+m+2]∗exp(−(WA)l+m+2[(WA)l+m+2]∗v)dv. |
The proof is completed utilizing Aⓦm,W=A(WA)l[(WA)l+m+2]†(WA)m.
Theorem 3.5. Let (A,W)∈Cp,n;k and m∈N. If H∈Cp×n, R(H)=R((AW)k), and N(H)=N([(WA)k]∗(WA)m), then
Aⓦm=∞∫0exp[−H(HWAWH)∗HWAWv]H(HWAWH)∗Hdv. |
Proof. Applying [37, Theorem 2.2], it follows that
Aⓦm,W=(WAW)(2)R((AW)k),N([(WA)k]∗(WA)m)=∞∫0exp[−H(HWAWH)∗HWAWv]H(HWAWH)∗Hdv, |
which completes the proof.
The W-m-WGI is applicable in studying solvability of some matrix and vector equations.
In the case that A∈Cm×n, x∈Cn, and b∈Cm, to find approximation solution to inconsistent system of linear equations Ax=b, a classical approach is to ask for, so called, generalized solutions, defined as solutions to GAx=Gb with respect to an appropriate matrix G∈Cn×m [38]. It is important to mention that the system GAx=Gb is consistent in the case rank(GA)=rank(G). Such approach has been exploited extensively. One particular choice is G=A∗, which leads to widely used least-squares solutions obtained as solutions to the normal equation A∗Ax=A∗b. Another important choice is m=n, G=Ak, and k=ind(A), which leads to the so called Drazin normal equation Ak+1x=Akb and usage of the Drazin inverse solution ADb.
Starting from the known equation WAWx=b, we use G=[(WA)k]∗(WA)m to obtain the following equation (4.1).
Theorem 4.1. If m∈N and (A,W)∈Cp,n;k, the general solution to
[(WA)k]∗(WA)m+1Wx=[(WA)k]∗(WA)mb,b∈Cn, | (4.1) |
is of the form
x=Aⓦm,Wb+(I−Aⓦm,WWAW)u, | (4.2) |
for arbitrary vector u∈Cn.
Proof. Let x be represented as in (4.2). Theorem 2.2 gives
[(WA)k]∗(WA)m+1WAⓦm,W=[(WA)k]∗(WA)m. |
So, x is a solution to (4.1) by
[(WA)k]∗(WA)m+1Wx=[(WA)k]∗(WA)m+1WAⓦm,Wb+[(WA)k]∗(WA)m+1W(I−Aⓦm,WWAW)u=[(WA)k]∗(WA)mb. |
If Eq (4.1) has a solution x, based on
Aⓦm,W=A[(WA)D]m+2(WA)k[(WA)k]†(WA)m, |
one concludes
Aⓦm,Wb=A[(WA)D]m+2(WA)k[(WA)k]†(WA)mb=A[(WA)D]m+2([(WA)k]†)∗[(WA)k]∗(WA)mb=A[(WA)D]m+2([(WA)k]†)∗[(WA)k]∗(WA)m+1Wx=A[(WA)D]m+2(WA)k[(WA)k]†(WA)m+1Wx=Aⓦm,WWAWx, |
which yields
x=Aⓦm,Wb+x−Aⓦm,WWAWx=Aⓦm,Wb+(I−Aⓦm,WWAW)x. |
Hence, x possesses the pattern (4.2).
Choosing m=2 or m≥k in Theorem 4.1, we obtain the next result.
Corollary 4.1. Let b∈Cn and (A,W)∈Cp,n;k.
(i) The general solution to
[(WA)k]∗(WA)3Wx=[(WA)k]∗(WA)2b | (4.3) |
possesses the form
x=Aⓦ2,Wb+(I−Aⓦ2,WWAW)u, |
for arbitrary u∈Cn.
(ii) If m≥k, the general solution to
(WA)m+1Wx=(WA)mb | (4.4) |
(or equivalently [(WA)k]∗(WA)m+1Wx=[(WA)k]∗(WA)mb)
possesses the form
x=AD,Wb+(I−AD,WWAW)u, |
for arbitrary u∈Cn.
We study assumptions which ensure the uniqueness of the solution to Eq (4.1).
Theorem 4.2. If m∈N and (A,W)∈Cp,n;k, x=Aⓦm,Wb is the unique solution to (4.1) in the space R((AW)k).
Proof. Theorem 4.1 implies that (4.1) has a solution x=Aⓦm,Wb∈R(Aⓦm,W)=R((AW)k).
For two solutions x,x1∈R((AW)k) to (4.1), by Lemma 2.2, we obtain
x−x1∈R((AW)k)∩N([(WA)k]∗(WA)m+1W)=R(Aⓦm,WWAW)∩N(Aⓦm,WWAW)={0}. |
Hence, the Eq (4.1) has uniquely determined solution x=Aⓦm,Wb in R((AW)k).
Theorem 4.2 gives the next particular results.
Corollary 4.2. Let b∈Cn and (A,W)∈Cp,n;k.
(i) x=Aⓦ2,Wb is the unique solution in R((AW)k) to (4.3).
(ii) x=AD,Wb is the unique solution in R((AW)k) to (4.4).
Recall that, by [39], for W∈Cn×p∖{0}, A∈Cp×n, ind(AW)=k1, ind(WA)=k2, and b∈R((WA)k2), x=AD,Wb is the uniquely determined solution to
WAWx=b,x∈R((AW)k1). |
Specifically, if A∈Cn×n, W=I, ind(A)=k, and b∈R(Ak), x=ADb is the unique solution to [40]
Ax=b,x∈R(Ak). |
For 1=ind(A) and b∈R(A), x=A#b is the uniquely determined solution to Ax=b. Notice that Theorem 4.2 and Corollary 4.2 recover the above mentioned results from [39] and [40].
The identity (resp., zero) ℓ×ℓ matrix will be denoted by Iℓ (resp., 0ℓ). Denote by Dpℓ, p≥1, the ℓ×ℓ matrix with its pth leading diagonal parallel filled by the entries of the vector 1={1,…,1}∈Cℓ−p and 0 in all other positions.
We perform numerical tests on the class of test matrices of index ℓ, given by
{(CIℓC1Iℓ0ℓC2Dpℓ), ℓ>0}, C,C1,C2∈C. | (5.1) |
Example 5.1. The test matrix A in this example is derived using ℓ=4 and C=2,C1=3,C2=1 from the test set (5.1), and W is derived using ℓ=4 and C=1,C1=3/2,C2=4 from the test set (5.1). Our intention is to perform numerical experiments on integer matrices using exact computation. Appropriate matrices are
A=(2I43I404D14)=(2000300002000300002000300002000300000100000000100000000100000000),W=(I43/2I4044D14)=(10003200001000320000100032000010003200000400000000400000000400000000). |
The matrices WA and AW fulfill k=ind(WA)=ind(AW)=2.
(a) In the first part of this example, we calculate the Drazin inverse, the core-EP inverse, and W-m-WGI class of inverses based on their definitions. The Drazin inverse of WA is computed using
(WA)D=(WA)2(WA5)†(WA)2=(1/2I43/4I4+3/8D14+3/2D24+3/4D340404)=(120003438323401200034383200120003438000120003400000000000000000000000000000000), |
and the core-EP inverse of WA is equal to
![]() |
The W-weighted Drazin inverse of A is equal to
AD,W=A[(WA)D]2=(WA)D |
and the W-weighted core-EP inverse of A is equal to
![]() |
The W-WGI (or W-1-WGI) inverse of A is given by
![]() |
the W-2-WGI inverse of A is equal to
![]() |
and for each m≥k the W-m-WGI inverse of A satisfies
![]() |
Additionally, Aⓦm=AD is checked for each m≥ind(A).
(b) Representations involved in Lemma 2.1 are verified using
![]() |
(c) In this part of the example, our goal is to verify results of Theorem 2.2.
(c1) The statements involved in Theorem 2.2(ⅳ) are verified as follows.
- In the case m=1 verification is confirmed by
AWAⓦ,WWAⓦ,W=Aⓦ,W;Aⓦ,W(WA)3W=(4I46I4+24D14+12D24+48D340404)=(4000624124804000624120040006240004000600000000000000000000000000000000)=(AW)2;[(WA)2]∗(WA)2WAⓦ,W=(8I412I4+6D140404)=(800012600080001260008000126000800012120001890061200945290246120362745291224612184527452)=[(WA)2]∗WA. |
- In the case m=2 results are confirmed by
AWAⓦ2,WWAⓦ2,W=Aⓦ2,W;Aⓦ2,W(WA)3W=(AW)2;[(WA)2]∗(WA)3WAⓦ2,W=2[(WA)2]∗WA=[(WA)2]∗(WA)2. |
- Representations in the case m≥3 are confirmed by
AWAⓦm,WWAⓦm,W=Aⓦm,W;Aⓦm,W(WA)3W=(AW)2;[(WA)2]∗(WA)m+1WAⓦm,W=m[(WA)2]∗WA=[(WA)2]∗(WA)m. |
(c2) The statements involved in Theorem 2.2(ⅵ) are verified using verification of part (ⅳ) and the following computation.
- In the case m=1
![]() |
- In the case m≥2 results are confirmed by
![]() |
Example 5.2. Consider A and W from Example 5.1 and the vector b=(22011201)T with intention to verify Theorem 4.1.
In the case m=1 of (4.1), the general solution to [(WA)2]∗(WA)2Wx=[(WA)2]∗WAb is equal to
x1=Aⓦ,Wb+(I−Aⓦ,WWAW)u, |
where u=(u1u2u3u4u5u6u7u8)T is a vector of unknown variables. Symbolic calculation gives
x1=(−3u52−6u6−3u7−12u8+52−3u62−6u7−3u8+52−3u72−6u8+3814(5−6u8)u5u6u7u8). |
Obtained vector x1 is verified using [(WA)2]∗(WA)2Wx1=[(WA)2]∗WAb=(404062060901594292).
In the case m≥2 of (4.1), the general solution to [(WA)2]∗(WA)m+1Wx=[(WA)2]∗(WA)mb is equal to
xm=Aⓦm,Wb+(I−Aⓦm,WWAW)u. |
Symbolic calculus produces
xm=(−3u52−6u6−3u7−12u8+134−3u62−6u7−3u8+4−3u72−6u8+3814(5−6u8)u5u6u7u8). |
Correctness of the vector xm is verified using
[(WA)2]∗(WA)m+1Wx2=[(WA)2]∗(WA)mb=(m−1)(1041281240156270426609). |
In this research, we present an extension of the m-weak group inverse (or m-WGI) on the set of rectangular matrices, called the W-weighted m-WGI (or W-m-WGI). The W-m-WGI class presents a new, wider class of generalized inverses since this class involves the m-WGI, W-weighted weak group, and W-weighted Drazin inverse as special cases. Various characterizations and representations of W-m-WGI are developed. Usability of the W-m-WGI class in solving some constrained and unconstrained matrix equations and linear systems is considered. Some new properties of the weighted generalized group inverse and some known properties of the W-weighted Drazin inverse are obtained as corollaries. The given numerical examples confirm the obtained results.
There is increasing interest in the investigation of the WGI and its generalizations, and so for further research in the near future, it may be interesting to consider its generalizations to Hilbert space operators or tensors, iterative methods for approximation of W-m-WGI, or recurrent neural network (RNN) models for computing W-m-WGI.
The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.
Dijana Mosić and Predrag Stanimirović are supported by the Ministry of Education, Science and Technological Development, Republic of Serbia, grant no. 451-03-65/2024-03/200124.
Predrag Stanimirović acknowledges support from the Science Fund of the Republic of Serbia, (No. 7750185, Quantitative Automata Models: Fundamental Problems and Applications - QUAM).
This work is supported by the Ministry of Science and Higher Education of the Russian Federation (Grant No. 075-15-2022-1121).
P. S. Stanimirović is an editorial board member for Electronic Research Archive and was not involved in the editorial review or the decision to publish this article. All authors declare that there are no competing interests.
[1] | A. Ben-Israel, T. N. E. Grevile, Generalized Inverses, Theory and Applications, 2nd edition, Canadian Mathematical Society, Springer, New York, Beflin, Heidelberg, Hong Kong, London, Milan, Paris, Tokyo, 2003. |
[2] |
R. E. Cline, T. N. E. Greville, A Drazin inverse for rectangular matrices, Linear Algebra Appl., 29 (1980), 53–62. https://doi.org/10.1016/0024-3795(80)90230-X doi: 10.1016/0024-3795(80)90230-X
![]() |
[3] |
K. M. Prasad, K. S. Mohana, Core-EP inverse, Linear Multilinear Algebra, 62 (2014), 792–802. https://doi.org/10.1080/03081087.2013.791690 doi: 10.1080/03081087.2013.791690
![]() |
[4] |
D. E. Ferreyra, F. E. Levis, N. Thome, Revisiting the core EP inverse and its extension to rectangular matrices, Quaestiones Math., 41 (2018), 265–281. https://doi.org/10.2989/16073606.2017.1377779 doi: 10.2989/16073606.2017.1377779
![]() |
[5] |
D. Mosić, Weighted core-EP inverse of an operator between Hilbert spaces, Linear Multilinear Algebra, 67 (2019), 278–298. https://doi.org/10.1080/03081087.2017.1418824 doi: 10.1080/03081087.2017.1418824
![]() |
[6] |
D. E. Ferreyra, F. E. Levis, N. Thome, Maximal classes of matrices determining generalized inverses, Appl. Math. Comput., 333 (2018), 42–52. https://doi.org/10.1016/j.amc.2018.03.102 doi: 10.1016/j.amc.2018.03.102
![]() |
[7] |
Y. Gao, J. Chen, Pseudo core inverses in rings with involution, Commun. Algebra, 46 (2018), 38–50. https://doi.org/10.1080/00927872.2016.1260729 doi: 10.1080/00927872.2016.1260729
![]() |
[8] |
M. Zhou, J. Chen, Integral representations of two generalized core inverses, Appl. Math. Comput., 333 (2018), 187–193. https://doi.org/10.1016/j.amc.2018.03.085 doi: 10.1016/j.amc.2018.03.085
![]() |
[9] |
R. Behera, G. Maharana, J. K. Sahoo, Further results on weighted core-EP inverse of matrices, Results Math., 75 (2020), 174. https://doi.org/10.1007/s00025-020-01296-z doi: 10.1007/s00025-020-01296-z
![]() |
[10] |
G. Dolinar, B. Kuzma, J. Marovt, B. Ungor, Properties of core-EP order in rings with involution, Front. Math. China, 14 (2019), 715–736. https://doi.org/10.1007/s11464-019-0782-8 doi: 10.1007/s11464-019-0782-8
![]() |
[11] |
I. Kyrchei, Determinantal representations of the core inverse and its generalizations with applications, J. Math., 2019 (2019). https://doi.org/10.1155/2019/1631979 doi: 10.1155/2019/1631979
![]() |
[12] |
H. Ma, P. S. Stanimirović, Characterizations, approximation and perturbations of the core-EP inverse, Appl. Math. Comput., 359 (2019), 404–417. https://doi.org/10.1016/j.amc.2019.04.071 doi: 10.1016/j.amc.2019.04.071
![]() |
[13] |
K. M. Prasad, M. D. Raj, Bordering method to compute core-EP inverse, Spec. Matrices, 6 (2018), 193–200. https://doi.org/10.1515/spma-2018-0016 doi: 10.1515/spma-2018-0016
![]() |
[14] |
M. M. Zhou, J. L. Chen, T. T. Li, D. G. Wang, Three limit representations of the core-EP inverse, Filomat, 32 (2018), 5887–5894. https://doi.org/10.2298/FIL1817887Z doi: 10.2298/FIL1817887Z
![]() |
[15] |
O. M. Baksalary, G. Trenkler, Core inverse of matrices, Linear Multilinear Algebra, 58 (2010), 681–697. https://doi.org/10.1080/03081080902778222 doi: 10.1080/03081080902778222
![]() |
[16] |
H. Wang, J. Chen, Weak group inverse, Open Math., 16 (2018), 1218–1232. https://doi.org/10.1515/math-2018-0100 doi: 10.1515/math-2018-0100
![]() |
[17] |
D. E. Ferreyra, V. Orquera, N. Thome, A weak group inverse for rectangular matrices, Rev. R. Acad. Cienc. Exactas Fˊis. Nat. Ser. A Mat., 113 (2019), 3727–3740. https://doi.org/10.1007/s13398-019-00674-9 doi: 10.1007/s13398-019-00674-9
![]() |
[18] |
D. Mosić, D. Zhang, Weighted weak group inverse for Hilbert space operators, Front. Math. China, 15 (2020), 709–726. https://doi.org/10.1007/s11464-020-0847-8 doi: 10.1007/s11464-020-0847-8
![]() |
[19] |
N. Liu, H. Wang, The characterizations of WG matrix and its generalized Cayley-Hamilton theorem, J. Math., 2021 (2021). https://doi.org/10.1155/2021/4952943 doi: 10.1155/2021/4952943
![]() |
[20] |
D. Mosić, P. S. Stanimirović, Representations for the weak group inverse, Appl. Math. Comput., 397 (2021), 125957. https://doi.org/10.1016/j.amc.2021.125957 doi: 10.1016/j.amc.2021.125957
![]() |
[21] |
H. Wang, X. Liu, The weak group matrix, Aequ. Math., 93 (2019), 1261–1273. https://doi.org/10.1007/s00010-019-00639-8 doi: 10.1007/s00010-019-00639-8
![]() |
[22] |
H. Yan, H. Wang, K. Zuo, Y. Chen, Further characterizations of the weak group inverse of matrices and the weak group matrix, AIMS Math., 6 (2021), 9322–9341. https://doi.org/10.3934/math.2021542 doi: 10.3934/math.2021542
![]() |
[23] |
M. Zhou, J. Chen, Y. Zhou, Weak group inverses in proper ∗-rings, J. Algebra Appl., 19 (2020), 2050238. https://doi.org/10.1142/S0219498820502382 doi: 10.1142/S0219498820502382
![]() |
[24] |
Y. Zhou, J. Chen, M. Zhou, m-weak group inverses in a ring with involution, Rev. R. Acad. Cienc. Exactas Fˊis. Nat. Ser. A Mat., 115 (2021). https://doi.org/10.1007/s13398-020-00932-1 doi: 10.1007/s13398-020-00932-1
![]() |
[25] |
W. Jiang, K. Zuo, Further characterizations of the m-weak group inverse of a complex matrix, AIMS Math., 7 (2022), 17369–17392. https://doi.org/10.1007/10.3934/math.2022957 doi: 10.1007/10.3934/math.2022957
![]() |
[26] |
D. E. Ferreyra, S. B. Malik, A generalization of the group inverse, Quaestiones Math., 46 (2023). https://doi.org/10.2989/16073606.2022.2144533 doi: 10.2989/16073606.2022.2144533
![]() |
[27] |
D. Mosić, P. S. Stanimirović, L. A. Kazakovtsev, Application of m-weak group inverse in solving optimization problems, Rev. R. Acad. Cienc. Exactas Fˊis. Nat. Ser. A Mat., 118 (2024), 13. https://doi.org/10.1007/s13398-023-01512-9 doi: 10.1007/s13398-023-01512-9
![]() |
[28] |
D. Mosić, D. Zhang, New representations and properties of m-weak group inverse, Results Math., 78 (2023). https://doi.org/10.1007/s00025-023-01878-7 doi: 10.1007/s00025-023-01878-7
![]() |
[29] |
V. Rakočević, Y. Wei, The representation and approximation of the W-weighted Drazin inverse of linear operators in Hilbert space, Appl. Math. Comput., 141 (2003), 455–470. https://doi.org/10.1016/S0096-3003(02)00267-9 doi: 10.1016/S0096-3003(02)00267-9
![]() |
[30] |
P. S. Stanimirović, V. N. Katsikis, H. Ma, Representations and properties of the W-Weighted Drazin inverse, Linear Multilinear Algebra, 65 (2017), 1080–1096. https://doi.org/10.1080/03081087.2016.1228810 doi: 10.1080/03081087.2016.1228810
![]() |
[31] |
M. P. Drazin, Weighted (b, c)-inverses in categories and semigroups, Commun. Algebra, 48 (2020), 1423–1438. https://doi.org/10.1080/00927872.2019.1687712 doi: 10.1080/00927872.2019.1687712
![]() |
[32] |
P. S. Stanimirović, D. Mosić, H. Ma, New classes of more general weighted outer inverses, Linear Multilinear Algebra, 70 (2022), 122–147. https://doi.org/10.1080/03081087.2020.1713712 doi: 10.1080/03081087.2020.1713712
![]() |
[33] |
M. P. Drazin, A class of outer generalized inverses, Linear Algebra Appl., 436 (2012), 1909–1923. https://doi.org/10.1016/j.laa.2011.09.004 doi: 10.1016/j.laa.2011.09.004
![]() |
[34] | C. W. Groetsch, Generalized inverses of linear operators: representation and approximation, in Monographs and Textbooks in Pure and Applied Mathematics, Marcel Dekker, Inc., New York, Basel, 37 (1977). |
[35] |
X. Liu, Y. Yu, J. Zhong, Y. Wei, Integral and limit representations of the outer inverse in Banach space, Linear Multilinear Algebra, 60 (2012), 333–347. https://doi.org/10.1080/03081087.2011.598154 doi: 10.1080/03081087.2011.598154
![]() |
[36] |
P. S. Stanimirović, Limit representations of generalized inverses and related methods, Appl. Math. Comput., 103 (1999), 51–68. https://doi.org/10.1016/S0096-3003(98)10048-6 doi: 10.1016/S0096-3003(98)10048-6
![]() |
[37] |
Y. Wei, D. S. Djordjević, On integral representation of the generalized inverse A(2)T,S, Appl. Math. Comput., 142 (2003), 189–194. https://doi.org/10.1016/S0096-3003(02)00296-5 doi: 10.1016/S0096-3003(02)00296-5
![]() |
[38] |
G. Maess, Projection methods solving rectangular systems of linear equations, J. Comput. Appl. Math., 24 (1988), 107–119. https://doi.org/10.1016/0377-0427(88)90346-9 doi: 10.1016/0377-0427(88)90346-9
![]() |
[39] |
Y. Wei, A characterization for the W-weighted Drazin inverse and a Cramer rule for the W-weighted Drazin inverse solution, Appl. Math. Comput., 125 (2002), 303–310. https://doi.org/10.1016/S0096-3003(00)00132-6 doi: 10.1016/S0096-3003(00)00132-6
![]() |
[40] | S. L. Campbell, C. D. Meyer, Generalized Inverses of Linear Transformations, Pitman, London, 1979. |
1. | Dijana Mosić, Predrag S. Stanimirović, Lev A. Kazakovtsev, Minimization problem solvable by weighted m-weak group inverse, 2024, 1598-5865, 10.1007/s12190-024-02215-z |