Diagonalization of Operator functions by algebraic methods

Matthias Stiefenhofer

REVIEW 2 major objections 6 minor 21 references

Diagonalization of Operator functions by algebraic methods

T0 review · 2 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Analytic operator families between Banach spaces admit a local block-diagonal normal form whose pole order is read off from stabilized Jordan chains.

desk verdict A constructive algebraic diagonalization result for analytic operator families, let down by a dense and typo-ridden analytic convergence section and an unnecessary editorial note about the paper's rejected predecessor. read the letter →

arxiv 2411.15905 v1 pith:VMWIACY5 submitted 2024-11-24 math.AG

Matthias Stiefenhofer This is my paper · ORCID

classification math.AG MSC 47A5615A2147A5547A53

keywords DiagonalizationJordanchainGeneralizedinverseToeplitzmatrixSmithformBanachspacesOperatorpowerseriesAnalyticfamily

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper establishes sufficient conditions under which an analytic family of bounded linear operators $L(\varepsilon)$ acting between real or complex Banach spaces can be locally diagonalized: there exist analytic near-identity transformations $\phi(\varepsilon)$ and $\psi(\varepsilon)$ for which $\psi^{-1}(\varepsilon)L(\varepsilon)\phi(\varepsilon)=\Delta(\varepsilon)$, a diagonal operator polynomial $\Delta(\varepsilon)=S_1P_1+\varepsilon S_2P_2+\cdots+\varepsilon^k S_{k+1}P_{k+1}$. The key hypothesis is stabilization of the Jordan chains at length $k$: no root element has finite rank greater than $k$, though root elements of infinite rank may exist. Together with closedness of the subspaces produced by an explicit recursion, this gives the diagonal form and, as a byproduct, a generalized inverse $L^{-1}(\varepsilon)$ whose pole at $\varepsilon=0$ has order exactly $k$. If correct, the result supplies the infinite-dimensional analogue of classical local normal forms and Smith factorization for analytic matrix functions, with explicit formulas for the transformations.

What carries the argument

The central object is the semi-infinite block matrix $M$ built by the recursion $S_{k+1}=(I-\mathcal P_1-\cdots-\mathcal P_k)[L_1\cdots L_k]M_k$, followed by splitting the current kernel and range to define the next subspaces and projections. Its columns are Jordan-chain vectors, and the identity $(L_0\cdots L_k)M^{(k+1)}=(S_1\cdots S_{k+1})$ converts the recursion into an actual triangularization of $L(\varepsilon)$. Once the Jordan chains stabilize at $k$, the part of $M$ below row $k+1$ becomes a Toeplitz matrix, meaning its entries are constant along diagonals, and row $k+1$ supplies the coefficients of the near-identity transformation $\phi(\varepsilon)$. A companion upper-triangular matrix $E$, whose inverse solves the triangular system defining the kernel of the Toeplitz coefficient matrix $\Delta_{k+1}$, makes the cancellation identities explicit.

What would settle it

For the 3x3 example in the paper, carry out the recursion one step beyond $k=3$ and check that the claimed period-4 pattern of $\phi_i$ and the displayed $\psi(\varepsilon)$ satisfy $\psi^{-1}L\phi=\Delta$; any disagreement in a coefficient would break the argument. In a general Banach space, take an analytic family stabilizing at $k=1$ whose kernel $N_1$ is closed but has no closed complement; if a bounded near-identity diagonalization with degree-1 $\Delta$ still exists, the closed-complement hypothesis is not necessary, while failure to construct the transformation would confirm that the hypothesis carries the argument.

Watch

Extended reading notes

Core claim

The central claim is Theorem 3(ii): under stabilization at $k$ and continuity of the projections $P_i$ and $\mathcal P_i$ defined along the recursion, the analytic family $L(\varepsilon)$ is diagonalized by analytic near-identity transformations according to $\psi^{-1}(\varepsilon)L(\varepsilon)\phi(\varepsilon)=\Delta(\varepsilon)$ with $\Delta(\varepsilon)=S_1P_1+\varepsilon S_2P_2+\cdots+\varepsilon^k S_{k+1}P_{k+1}$. The proof is constructive and algebraic: a recursion splits the domain and range into direct sums $B=N_1^c\oplus\cdots\oplus N_{k+1}^c\oplus N_{k+1}$ and $\tilde B=R_1\oplus\cdots\oplus R_{k+1}\oplus R_{k+1}^c$, and the columns of a semi-infinite block matrix $M$ record Jordan chains and, once stabilization occurs, become Toeplitz. Reading row $k+1$ of $M$ gives the coefficients of $\phi$, while $\psi$ is assembled from the operators $S_i$; the same data yield a generalized inverse with pole order $k$, explicit continuation of kernels and ranges, and a Smith factorization.

Load-bearing premise

The construction requires that at every step of the recursion the chosen algebraic complements $N_i^c$ and $R_i^c$ be closed subspaces; in general Banach spaces closed complements need not exist, so this is an extra assumption and not a consequence of stabilization.

Editorial extensions

If this is right

For $\varepsilon\neq0$, $L^{-1}(\varepsilon)=\phi(\varepsilon)\Delta^{-1}(\varepsilon)\psi^{-1}(\varepsilon)$ is a generalized inverse that is analytic in a punctured neighbourhood of $0$ and has a pole of order exactly $k$ at $\varepsilon=0$.
Kernels and ranges continue analytically to the limit spaces: $N[L(\varepsilon)]=\phi(\varepsilon)N_{k+1}$ and $R[L(\varepsilon)]=\psi(\varepsilon)(R_1\oplus\cdots\oplus R_{k+1})$ for $\varepsilon\neq0$, so no jump in dimension occurs as $\varepsilon$ passes through $0$.
The diagonal polynomial factorizes as $\Delta(\varepsilon)=S_PP(\varepsilon)$, yielding a Smith factorization $L(\varepsilon)=\mathcal A(\varepsilon)P(\varepsilon)\phi^{-1}(\varepsilon)$ and a version in which $L(\varepsilon)$ is reduced to the constant operator $S_P$ after inverting $P$.
A meromorphic operator family $M(\varepsilon)$ with pole order $p$ is covered by applying the theorem to the analytic family $\varepsilon^pM(\varepsilon)$, so the diagonalization and pole-order statements extend to the meromorphic setting.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The explicit Toeplitz structure means that in finite-dimensional problems the recursion is an algorithm for computing the diagonal form, the transformation coefficients, and the pole order by linear algebra alone; complexity and numerical conditioning are not discussed in the paper.
Since the recursion is purely formal before analyticity is invoked, the same proof should survive with the field $\mathbb K$ replaced by a suitable ring or with convergence measured in a different topology, a direction the paper only hints at.
The recursion's reliance on chosen complements suggests that in non-Hilbert spaces the expected failure mode is the absence of closed ranges rather than the absence of stabilization; a concrete family with a non-complemented kernel would test where the theorem's boundary lies.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The paper delivers a genuinely constructive algebraic recipe for putting an analytic operator family L(epsilon) into local diagonal normal form, under Jordan-chain stabilization at length k plus continuity of the associated projections. The fine direct-sum resolution (1.3) and the explicit formulas for the near-identity transformations phi and psi go beyond the generalized-inverse existence results in [5,6,11]; that is the real contribution. The algebraic recursion in Section 3.1 is clearly stated, Lemma 1 and Theorems 1-2 look sound, and the 3x3 example checks out. The proof is self-contained and there are no fitted parameters or circular steps.

Now the soft spots, in proportion. First, the load-bearing hypothesis is the closedness of the complements N_i^c and R_i^c at every stage, equivalently the continuity of the projections P_i and P-tilde_i (Remark 1). The stress-test concern is valid: stabilization at k does not guarantee that such closed complements exist in infinite-dimensional Banach spaces (c0 inside l_infinity is the standard counterexample). The proof needs the bounded inverse theorem to get bounded S_i^{-1}, and it needs the finite data M(2k+1) to be bounded to set up the fixed-point equation (5.15)-(5.16). So the analytic conclusion is genuinely conditional on an extra assumption. To the paper's credit, the assumption is stated explicitly rather than smuggled, but the paper should discuss cases where stabilization holds and closed complements fail.

Second, Section 5 is the weakest part. The fixed-point equation is long, the text around (4.23), (5.10), and (3.8) is corrupted by typos, and the convergence proof is not verified line by line. A referee would need a substantial rewrite of that section.

Third, the manuscript opens with a self-referential note saying it is a revised and rejected version of [19] and that no substantial gap was identified. That editorial baggage does not belong in a submission and it complicates the novelty claim: the core diagonalization already appears in the author's own [19]. The incremental content here is the explicit transformation and the fine resolution; that should be stated in the introduction without the rejection narrative.

Who is this for? Researchers in spectral theory, bifurcation, and operator-function theory who want explicit normal forms and generalized inverses. The algebraic part is solid; the analytic part needs work. I would not desk-reject it: the construction is of interest and the algebraic proof is careful enough to merit a serious referee. But I would send it back with major revision, requiring removal of the editorial note and a clean, verifiable rewrite of Section 5.

Referee Report

2 major / 6 minor

Summary. The paper develops an algebraic recursion for block-diagonalizing an analytic family L(ε)=Σ ε^i L_i of bounded operators between Banach spaces. Under the assumption that the Jordan chains stabilize at length k and that the recursively chosen algebraic complements are closed (equivalently, the associated projections are continuous), it constructs formal power series φ and ψ and a diagonal operator polynomial Δ(ε)=S_1P_1+...+ε^k S_{k+1}P_{k+1} satisfying L(ε)=ψ(ε)Δ(ε)φ^{-1}(ε). Theorems 1 and 2 supply the formal triangularization and factorization; Theorem 3 asserts analyticity of the transformations via a fixed-point equation in (5.15)-(5.16), and derives a generalized inverse with pole order k plus smooth continuation of kernels and ranges. A 3x3 matrix example illustrates the recursion and exhibits periodic coefficients.

Significance. If the claims are correct, the paper gives a constructive, self-contained route to a local Smith-type normal form for analytic operator families in infinite-dimensional Banach spaces under explicit hypotheses. The Toeplitz/M-matrix perspective and the reduction of the convergence question to a finite-data fixed-point equation are genuinely useful, and the worked example makes the recursion concrete and checkable. The main theorem is honestly conditional on closed complements at every stage, and the paper does not overstate its scope. Prior work by Bart-Kaashoek-Lay and Kaballo established smooth generalized inverses under similar stabilization conditions; the extra diagonal polynomial structure obtained here is a real strengthening. The formal parts of the recursion are presented in a way that can be verified independently, which is a strength of the paper.

major comments (2)

[Section 5, Eqs. (5.15)-(5.16)] The proof of Theorem 3 rests on the vector fixed-point equation ā(ε)=q̄(ε)+ε Q(ε)ā(ε). As written, the derivation is too compressed. In (5.14)-(5.15), the bracketed expressions involve block rows with 'k times' zero entries and several index shifts whose meaning is not fully specified, and the claim that q̄(ε) and Q(ε) are analytic in the operator norm and depend only on the finite data M(2k+1) together with the series in (5.17) is asserted rather than demonstrated. Since this is the only mechanism by which the formal power series φ(ε) of (4.24) is shown to converge, the paper should give a complete derivation of the equation, explicit formulas for q̄ and Q, a proof that I-εQ(ε) is invertible for small ε, and a convergence argument for the solution ā. This is a local fix, but it is load-bearing for Theorem 3(i)-(ii).
[Theorem 3 and Remark 1] The central hypothesis of Theorem 3 is continuity of the projections P_i and 𝒫_i, i.e. closedness of the complements N_i^c and R_i^c at each stage. Stabilization at k does not imply that such closed complements exist in arbitrary Banach spaces; indeed, a closed subspace may have only non-closed algebraic complements. The paper acknowledges this in Remark 1, but the statements in the Introduction and Theorem 3 should make clear that this is an additional, independent assumption rather than a consequence of stabilization. Relatedly, Remark 3's assertion that a Fredholm L0 automatically yields continuous projections needs a short proof: because N(L0) is finite-dimensional and R(L0) is finite-codimensional, complements can be chosen closed, and the same remains true inductively. Without such a justification, a reader cannot tell how restrictive the theorem is.

minor comments (6)

[Title page / Section 1] The opening line of the manuscript ('The paper represents a revised and rejected version of [19]...') is inappropriate for a journal submission and should be deleted or rewritten in a neutral way.
[Eq. (3.7)] The summation notation in the formula for E_{i,k+1} is garbled; it should read E_{i,k+1} = -S_i^{-1} 𝒫_i Σ_{ν=i+1}^{k+1} S̅_ν E_{ν,k+1} to be unambiguous.
[Section 4, around (4.23)] The block entries M_{i,j} are used throughout the M-matrix before the notation is defined; please add a sentence stating that M_{i,j} are bounded operators from B to B when the matrix is introduced.
[Theorem 3 proof] The invertibility of ψ(ε) and φ(ε) for small ε is used implicitly. Since these transformations are analytic and equal I at 0, this follows from the Neumann series in L(B,B) and L(B̃,B̃); state this explicitly.
[Remark 1] The phrasing 'closed complements N_i^c and R_i^c' should be read together with the equivalence 'projection continuous iff both the image and kernel are closed'; making that equivalence explicit in Remark 1 would prevent a common misreading.
[References] Several items in the reference list are marked 'not submitted'; if they are only preprints or unpublished notes, this should be stated clearly or the citations should be reduced.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: Theorem 3 is a constructive proof under explicit hypotheses; self-citations are contextual, not load-bearing.

full rationale

I walked the paper's derivation chain from the recursion (3.5)-(3.8) through Lemma 1, Theorem 1, Theorem 2, and Theorem 3. The diagonalization claim psi^{-1}(epsilon)L(epsilon)phi(epsilon)=Delta(epsilon) is not obtained by assuming the conclusion: phi(epsilon) is defined from row k+1 of the M-matrix in (4.24), psi(epsilon) is defined explicitly in (4.34), and the factorization S(epsilon)=psi(epsilon)Delta(epsilon) is verified by a Cauchy-product computation using the algebraic properties (4.21) and (4.32) that are themselves proved from the recursion. No fitted parameters are used, and no quantity is predicted from a subset of data. The continuity/closedness hypotheses in Theorem 3 are explicit assumptions: the proof invokes the bounded inverse theorem only after noting that continuity of the projections implies closedness of the relevant subspaces. This is a conditional hypothesis, not a circularly imported consequence. The paper does cite the author's earlier work heavily, and it states that it is a revised version of [19], but the central recursion is restated and proved in the present paper, and the external invocation is Taylor's bounded inverse theorem [20], which is standard and independent. Therefore, although the self-citation density is high, none of the cited prior work carries the load of the diagonalization proof. I found no step where an equation reduces to its own input by definition, no fitted parameter renamed as a prediction, and no uniqueness or ansatz smuggled in via self-citation. Score 2 reflects the heavy but non-load-bearing self-citation, not actual circularity.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no fitted parameters and no new physical or mathematical entities. It rests on the stated structural assumptions (analyticity, stabilization, closedness of complements) and standard functional-analytic facts. The main theorem is conditional, not a parameter-free derivation.

assumptions (5)

domain assumption The operator family L(epsilon)=sum epsilon^i L_i is analytic in a neighborhood of 0 with bounded operators L_i in L[B,B-tilde].
Used throughout; Theorem 3 adds analyticity to ensure convergence of the formal power series solutions (Section 5).
domain assumption Jordan chains stabilize at level k: rank(b0) in {0,...,k} union {infinity} for all b0 in B, and some b0 has rank exactly k.
This is the central structural condition that forces the Toeplitz repetition in the M-matrix (Definition 1(ii), Lemma 1(iii)).
domain assumption At each recursion stage, algebraic complements N_i^c and R_i^c exist and can be chosen so that the associated projections P_i and P-tilde_i are continuous, equivalently the subspaces are closed.
Needed for boundedness of the inverse operators S_i^{-1} and for analyticity of phi and psi in Theorem 3; not automatic in general Banach spaces.
standard math Bounded inverse theorem (open mapping/closed graph) for Banach spaces.
Invoked in Section 5 to conclude S_i^{-1} is bounded when the projections are continuous.
standard math Existence of algebraic complements in vector spaces via Zorn/choice.
Used to build the direct-sum decompositions (1.3) before closedness is imposed; see Section 3.1.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Diagonalization of Operator functions by algebraic methods." pith.science (2026). https://pith.science/paper/VMWIACY5

@misc{pith2026241115905,
  author       = {Pith},
  title        = {Pith review of: Diagonalization of Operator functions by algebraic methods},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VMWIACY5}},
  note         = {Machine review of arXiv:2411.15905}
}

read the original abstract

We give conditions for local diagonalization of an analytic operator family to a diagonal operator polynomial. The families are acting between real or complex Banach spaces. The basic assumption is given by stabilization of the Jordan chains at length k in the sense that no root elements with finite rank above k are allowed to exist. Jordan chains with infinite rank may appear. Decompositions of the linear spaces are constructed with corresponding subspaces assumed to be closed. These assumptions ensure finite pole order equal to k of the generalized inverse. The Smith form and smooth continuation of kernels and ranges to appropriate limit spaces arise immediately. An algebraically oriented and self-contained approach is used, based on a recursion that allows for construction of power series solutions. The power series solutions are convergent, as soon as analyticity and continuity of related projections are assumed.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 21 canonical work pages

[19]

Stiefenhofer, Formal and Analytic Diagonalization of Operator functions , arXiv:2305

M. Stiefenhofer, Formal and Analytic Diagonalization of Operator functions , arXiv:2305. 14228, [math. AG], 2023

work page 2023
[1]

Albrecht, P

A. Albrecht, P. Howlett, C. Pearce, The fundamental equations for inversion of operator pencils on Banach space, Journal of Mathematical Analysis and Applications, 413, 411–421, 2014

work page 2014
[2]

Albrecht, P

A. Albrecht, P. Howlett, G. Verma, Inversion of operator pencils on Banach space using Jordan chains when the generalized resolvent has an isolated essential singularity , Linear Algebra and its Applications, 595, 33–62, 2020. 27

work page 2020
[3]

Howlett, K

P. Howlett, K. Avrachenkov, C. Pearce, V. Ejov, Inversion of analytically perturbed linear op- erators that are singular at the origin , Journal of Mathematical Analysis and Applications, 353 (1), 68-84, 2009

work page 2009
[4]

Albrecht, P

A. Albrecht, P. Howlett, G. Verma, Inversion of operator pencils on Hilbert space , Journal of the Australian Mathematical Society, 108, 145–176, 2018

work page 2018
[5]

Bart, M.A

H. Bart, M.A. Kaashoek, D.C. Lay, Stability properties of finite meromorphic operator functions I, II, III, Indagationes Mathematicae, Volume 77, Issue 3, 217-259, 1974

work page 1974
[6]

H. Bart, W. Kaballo, Local invertibility of meromorphic operator functions, Proceedings of the Royal Irish Academy, Section A, 78, 37-50, 1978

work page 1978
[7]

Beyn, An integral method for solving nonlinear eigenvalue problems, Linear Algebra and Its Applications 436, 3839–3863, 2012

W.-J. Beyn, An integral method for solving nonlinear eigenvalue problems, Linear Algebra and Its Applications 436, 3839–3863, 2012

work page 2012

Show all 21 references

[8]

Esquinas, Optimal multiplicity in local bifur cation theory, II: General case, J

J. Esquinas, Optimal multiplicity in local bifur cation theory, II: General case, J. Diff. Equ. 75, 206-215, 1988

1988
[9]

Gohberg, J

I. Gohberg, J. Leiterer, Holomorphic Operator Functions of One Variable and Applications, Op- erator Theory 192, ISBN 978-3-0346-0125-2, Birkhäuser, 2009

2009
[10]

Gohberg, M.A

I. Gohberg, M.A. Kaashoek, F. van Schagen, On the local theory of regular analytic matrix func- tions, Linear Algebra and its Applications, Volume 182, 9-25, 1993

1993
[11]

Kaballo, Meromorphic generalized inverses of operator functions, Indagationes Mathemat- icae 23, 970-994, Elsevier, 2012

W. Kaballo, Meromorphic generalized inverses of operator functions, Indagationes Mathemat- icae 23, 970-994, Elsevier, 2012

2012
[12]

López-Gómez, C

J. López-Gómez, C. Mora-Corral, Algebraic Multiplicity of Eigenvalues of Linear Operators, Op- erator Theory 177, ISBN 978-3-7643-8400-5, Birkhäuser, 2007

2007
[13]

Mennicken, M

R. Mennicken, M. Möller, Non-Self-Adjoint Boundary Eigenvalue Problems , North -Holland Mathematics Studies, vol. 192, North-Holland Publishing Co., Amsterdam, 2003

2003
[14]

Stiefenhofer, Singular Perturbation and Bifurcation in case of Dictyostelium discoideum, University of Konstanz, Research Gate: M

M. Stiefenhofer, Singular Perturbation and Bifurcation in case of Dictyostelium discoideum, University of Konstanz, Research Gate: M. Stiefenhofer , Hartung -Gorre-Verlag Konstanz, 1995

1995
[15]

Stiefenhofer, Linearization and Lemma of Newton for Operator functions , arXiv:2309

M. Stiefenhofer, Linearization and Lemma of Newton for Operator functions , arXiv:2309. 15548v2 [math.FA], 2024, submitted

2024
[16]

Stiefenhofer, A note on an expansion formula with application to nonlinear differential al- gebraic equations, arXiv:1912.10735 [math.AG], 2019, not submitted

M. Stiefenhofer, A note on an expansion formula with application to nonlinear differential al- gebraic equations, arXiv:1912.10735 [math.AG], 2019, not submitted

1912 arXiv
[17]

Stiefenhofer, Direct sum condition and Artin Approximation in Banach spaces, arXiv:1905

M. Stiefenhofer, Direct sum condition and Artin Approximation in Banach spaces, arXiv:1905. 07583v1 [math.AG], 2019, not submitted

1905
[18]

Stiefenhofer, Fine Resolution of 𝑘𝑘-transversal Cones, arXiv:2004.08615, [math

M. Stiefenhofer, Fine Resolution of 𝑘𝑘-transversal Cones, arXiv:2004.08615, [math. AG], 2020, not submitted

2004 arXiv
[20]

Taylor, Introduction to Functional Analysis, John Wiley & Sons, 1967

A.E. Taylor, Introduction to Functional Analysis, John Wiley & Sons, 1967

1967
[21]

Töplitz, Zur Theorie der quadratischen und bilinearen Formen von unendlichvielen Verän- derlichen, Mathematische Annalen 70, 351-376, 1911

O. Töplitz, Zur Theorie der quadratischen und bilinearen Formen von unendlichvielen Verän- derlichen, Mathematische Annalen 70, 351-376, 1911

1911

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [19]

Stiefenhofer, Formal and Analytic Diagonalization of Operator functions , arXiv:2305

M. Stiefenhofer, Formal and Analytic Diagonalization of Operator functions , arXiv:2305. 14228, [math. AG], 2023

work page 2023

[2] [1]

Albrecht, P

A. Albrecht, P. Howlett, C. Pearce, The fundamental equations for inversion of operator pencils on Banach space, Journal of Mathematical Analysis and Applications, 413, 411–421, 2014

work page 2014

[3] [2]

Albrecht, P

A. Albrecht, P. Howlett, G. Verma, Inversion of operator pencils on Banach space using Jordan chains when the generalized resolvent has an isolated essential singularity , Linear Algebra and its Applications, 595, 33–62, 2020. 27

work page 2020

[4] [3]

Howlett, K

P. Howlett, K. Avrachenkov, C. Pearce, V. Ejov, Inversion of analytically perturbed linear op- erators that are singular at the origin , Journal of Mathematical Analysis and Applications, 353 (1), 68-84, 2009

work page 2009

[5] [4]

Albrecht, P

A. Albrecht, P. Howlett, G. Verma, Inversion of operator pencils on Hilbert space , Journal of the Australian Mathematical Society, 108, 145–176, 2018

work page 2018

[6] [5]

Bart, M.A

H. Bart, M.A. Kaashoek, D.C. Lay, Stability properties of finite meromorphic operator functions I, II, III, Indagationes Mathematicae, Volume 77, Issue 3, 217-259, 1974

work page 1974

[7] [6]

H. Bart, W. Kaballo, Local invertibility of meromorphic operator functions, Proceedings of the Royal Irish Academy, Section A, 78, 37-50, 1978

work page 1978

[8] [7]

Beyn, An integral method for solving nonlinear eigenvalue problems, Linear Algebra and Its Applications 436, 3839–3863, 2012

W.-J. Beyn, An integral method for solving nonlinear eigenvalue problems, Linear Algebra and Its Applications 436, 3839–3863, 2012

work page 2012