Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Accuracy and componentwise accuracy in multilinear PageRank

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read For alpha <= 1/2, multilinear PageRank solutions admit componentwise perturbation bounds that do not blow up at the alpha = 1/2 singularity.

desk verdict Worth refereeing: solid componentwise perturbation theory and effective GTH-based algorithms, but the headline 'no blow-up at α=1/2' claim rests on an unstated irreducibility condition. read the letter →

arxiv 2506.18356 v1 pith:VMEE4JNU submitted 2025-06-23 math.NA cs.NA

Mehdi Najafi Kalyani , Federico Poloni This is my paper · ORCID
classification math.NAcs.NA MSC 15A6915B5165H10
keywords multilinearPageRankcomponentwiseperturbationboundpartialinverseofM-matricesNewtonmethodGTHalgorithmblockJacobiquadraticvectorequationsstochastictensors
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 studies how accurately solutions of the multilinear PageRank equation $x=(1-\alpha)v+\alpha \mathcal{P}x^2$ can be computed and how stable they are under small data perturbations. The central claim is that standard error bounds based on the norm of the Jacobian are pessimistic: for $\alpha\le 1/2$, the minimal solution obeys a componentwise perturbation bound that stays finite as $\alpha\to 1/2$, the point where the Jacobian becomes singular, as long as perturbations preserve the stochastic normalization. For $\alpha>1/2$, the limiting accuracy of Newton's method is controlled by the smaller quantity $u\|\lvert R_{x^*}^{-1}\rvert x^*\|$ rather than by $u\|R_{x^*}^{-1}\|\|x^*\|$. The paper also proposes subtraction-free GTH-based variants of Newton and block-Jacobi iterations that compute the minimal solution with componentwise accuracy even when the Jacobian is nearly singular. A new partial inverse of an M-matrix, which remains bounded as the matrix approaches singularity, is the mechanism behind this improved bound.

What carries the argument

The central object is a decomposition $M^{-1}=R+S$ for an M-matrix, obtained from the all-minors matrix tree theorem: $R$ is a rank-one matrix, and $S$ is a partial inverse that acts like the inverse on zero-sum vectors and remains bounded as the row sums of $M$ tend to zero (Theorem 3.4). Applied to $R_m^T$, this $S$ replaces the exploding $R_m^{-1}$ in the perturbation analysis, giving the finite quantity $\omega$ in (4.8). The computational half uses a triplet representation of the Newton or block-Jacobi linear system and the GTH factorization, together with the subtraction-free residual update $r=\alpha \mathcal{P}h^2$; these avoid cancellation so that small entries of $m$ are computed with relative accuracy.

What would settle it

Take a tensor with two disconnected blocks, set $\alpha$ extremely close to $1/2$, perturb $\mathcal{P}$ and $v$ by a zero-sum componentwise perturbation of size $\varepsilon$, and compare the observed $d(\tilde m,m)$ with $2\omega\gamma\varepsilon$: if $\omega$ diverges or the inequality fails, the irreducibility support condition is load-bearing.

Watch

Extended reading notes

Core claim

The paper establishes two accuracy results. First, for the multilinear PageRank equation with $\alpha\le 1/2$, the minimal solution $m$ obeys the componentwise perturbation bound $d(\tilde m,m)\le 2\omega\gamma\varepsilon$ for perturbations that preserve the stochastic constraints $1^T a=1-\alpha$ and $1^T \mathcal{P}_{(1)}=\alpha 1^T_{n^2}$, where $d$ is the componentwise distance $d(\tilde x,x)=\max_i |\tilde x_i-x_i|/|x_i|$; here $\omega$ is defined through a bounded partial inverse $S$ of $R_m^T$ and does not diverge as $\alpha\to 1/2$, unlike $R_m^{-1}$. Second, for $\alpha>1/2$, Newton's limiting accuracy is $u\|\lvert R_{x^*}^{-1}\rvert x^*\|+o(u)$, replacing the classical $u\|R_{x^*}^{-1}\|\|x^*\|$; numerical experiments confirm that the smaller quantity tracks the observed stagnation. The same analysis yields GTH-based subtraction-free Newton and block-Jacobi algorithms that compute the minimal solution with componentwise accuracy near the singularity.

Load-bearing premise

The proof assumes the tensor's underlying graph stays strongly connected near $\alpha=1/2$ (so the partial inverse $S$ stays bounded) and that the minimal solution has no zero entries; if the graph becomes reducible, the claimed finite bound can become vacuous.

Editorial extensions

If this is right

  • Near $\alpha=1/2$, the minimal solution can be computed with componentwise accuracy by the Newton-GTH and block-Jacobi-GTH algorithms even though the Jacobian is ill-conditioned.
  • Under stochastic-normalization-preserving perturbations, errors on very small entries of the minimal solution are not amplified by the condition number of $R_m$.
  • For $\alpha>1/2$, the stagnation level of Newton's method is predicted by $u\|\lvert R_{x^*}^{-1}\rvert x^*\|$, so small solution entries do not necessarily pollute the computation as much as a normwise analysis suggests.
  • The partial-inverse argument applies to general quadratic vector equations of the form $x=a+Bx^2$ with $\rho(Bm:+B:m)<1$, giving componentwise bounds beyond the PageRank setting.

Reading between the lines

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

  • Editorial inference: the decomposition $M^{-1}=R+S$ is effectively a pseudoinverse on the zero-sum subspace, so it could produce condition-number-free componentwise bounds for other stochastic-structure problems, such as ordinary PageRank or Perron vectors of column-stochastic matrices, whenever data are renormalized after perturbation.
  • Editorial inference: the paper's experiments note that the perturbation bound seems to hold for stochastic solutions with $\alpha>1/2$ even though the proof covers minimal solutions; a targeted test in this regime could turn that numerical observation into a theorem.
  • Editorial inference: a practical takeaway for implementers is to stop Newton-type iterations on a componentwise residual, since the subtraction-free residual $r=\alpha \mathcal{P}h^2$ makes the relative error in small entries the natural cost function.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper studies componentwise perturbation stability and numerical accuracy for the multilinear PageRank equation x = (1-α)v + αP x^2, and the more general quadratic equation x = a + Bx^2. The main theoretical results are: (i) a componentwise perturbation bound for the minimal solution when α < 1/2 (Theorem 4.3); (ii) an improved bound under zero-sum stochasticity-preserving perturbations, based on a decomposition of the inverse of an M-matrix into a rank-one part and a bounded remainder S, giving d(m̃,m) ≤ 2ωγε with ω defined via S (Theorem 4.4); (iii) an analysis of the limiting accuracy of Newton's method for α > 1/2, replacing the classical factor ‖R^{-1}_{x*}‖‖x*‖ by ‖|R^{-1}_{x*}|x*‖ (Theorems 6.1 and 6.2); and (iv) subtraction-free GTH-based Newton and block-Jacobi algorithms for computing the minimal solution with componentwise accuracy. Numerical experiments on real and synthetic tensors compare the proposed algorithms with standard Newton and illustrate the bounds.

Significance. If the central claims hold, the paper gives a genuinely useful message for multilinear PageRank: componentwise stability can be much better than classical Jacobian-based bounds suggest, both for perturbations and for rounding errors in Newton-type iterations. The combinatorial matrix-tree decomposition of the inverse into a bounded remainder S and a rank-one part is an interesting technical contribution that may have applications beyond PageRank. The algorithms are concrete, and the paper explicitly identifies the novel subtraction-free residual formula r = αPh². The theoretical derivations are internally consistent, and I checked the decomposition of Theorem 3.4 on small examples. The main caveat is that the headline 'ignoring the α = 1/2 singularity' claim depends on a boundedness property of S that is proved only under hypotheses not stated or verified in the main application, so the central claim needs an additional assumption or a separate argument.

major comments (3)
  1. [Section 4.3, Theorems 3.4 and 4.4] The boundedness of the remainder S is the load-bearing ingredient in the claim that the componentwise bound d(m̃,m) ≤ 2ωγε 'ignores the singularity' at α = 1/2. Theorem 3.4 proves boundedness only for sequences M(w_k) of irreducible matrices with constant off-diagonal entries while w_k → 0. In Section 4.3, M = R_m^T has w = (1-2α)1 → 0, but the off-diagonal entries of R_m^T change with α through m = m(α), and Theorem 4.4 neither assumes nor verifies that R_m^T is irreducible. If the support graph of R_m^T is reducible, the polynomials D_i in the proof of Theorem 3.4 can vanish, the bound on S may fail, and ω defined in (4.8) can diverge as α → 1/2, e.g. like 1/(1-2α). In that case the formal inequality d(m̃,m) ≤ 2ωγε is still true for each fixed α, but it blows up exactly at the singularity that the abstract promises to ignore. The theorem needs an explicit irreducibility/positive-support assumption on P or R_m^T, or a separate argument showing that the tensor stochasticity structure keeps S bounded even when R_m^T is reducible; neither is supplied.
  2. [Section 4.3, proof of Theorem 4.4] The fixed-point argument defines Bkh = {h : 1^T h = 0, |h| ≤ k_h m} and claims φ(Bkh) ⊆ Bkh, where φ(h) = S̃^T(ã - a + (B̃ - B)m² + B̃h²). The proof bounds |φ(h)| componentwise but does not show 1^T φ(h) = 0. This zero-sum preservation is necessary for the Brouwer fixed-point argument to produce a point in Bkh. It can be supplied from the conditions 1^T(ã-a)=0, 1^T(B̃-B)(1)=0, 1^T B̃ h² = α(1^T h)² = 0, and 1^T R̃_m = (1-2α)1^T; the argument should be written out.
  3. [Section 7, Examples 4 and 5] The numerical experiments include stochastic tensors with many zero entries, such as the 4×17 tensor in Example 4 and the tensor in Example 5. These are exactly the cases where the missing irreducibility hypothesis in Theorem 4.4 is most relevant. Before accepting the experimental confirmation of the non-blow-up bound as evidence for the theorem, the authors should verify that R_m^T is irreducible (or that the boundedness of S holds) for these examples, or state which additional property of these tensors makes the application of Theorem 3.4 valid.
minor comments (5)
  1. [Section 2.3 and Theorems 4.3-4.4] The assumption m > 0 is stated in Section 2.3, but it should be recalled explicitly in the statements of Theorems 4.3 and 4.4 and in the abstract-level claims, since the componentwise distance d(·,·) divides by m_i and the quantities κ,ω in (4.2),(4.8) are undefined when some m_i = 0.
  2. [Theorem 6.2, equation (6.1)] The estimate |e_k| ≤ γ̃_{n²} x* + o(u) mixes the limit k → ∞, used to justify x̃_k - x* = o(1), with the limit u → 0 used in the o(u) term. The asymptotic regime in which (6.1) is intended should be stated more precisely (e.g., u → 0 with k large enough so that x̃_k is within a fixed contraction region).
  3. [Section 6, paragraph before Theorem 6.1] The notation ∥R^{-1}_{x*} e∥ in Theorem 6.1 is used for a vector norm applied to a vector, while ∥R^{-1}_{x*}∥ is an induced matrix norm; the norms should be named explicitly to avoid confusion between the two inequalities.
  4. [Section 5.2, variant of block Jacobi-GTH] The text correctly notes that the variant with the Newton-based scalar z_k may violate w_k ≤ m and has no monotonic convergence guarantee. Since the algorithm is still recommended on the basis of experiments, it would help to report a convergence failure rate or a theoretical condition under which the iteration converges.
  5. [Section 7, Figures 5-7] The paper observes that the theoretical perturbation bounds are quite loose. A short explanation of the main sources of looseness (the factor γ from Theorem 3.2, the use of ω or κ, and the quadratic term in the fixed-point argument) would make the experimental section more informative.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the perturbation bounds are a priori data-dependent constants, and the self-citations are background results rather than load-bearing target assumptions.

full rationale

The derivation chain is self-contained in the relevant sense. Theorem 4.4 proves the componentwise bound d(\tilde m,m) \le 2\omega\gamma\varepsilon with \omega defined in (4.8) from z=S^T m, where S is the bounded part of R_m^{-T} from Theorem 3.4. The constant \omega is computed from the exact minimal solution and the matrix R_m; it is not fitted to the observed errors. Section 7 compares the theoretical bound with independently computed componentwise errors in Figures 5 and 6, and does not use those errors to tune \omega, \gamma, or \varepsilon. The self-citations are background: [16] is cited for existence, minimality, and monotone convergence of the minimal solution (Theorem 2.1), and [13] is cited for known behavior of stochastic solutions for \alpha>1/2 and for a Perron-vector-based algorithm. Neither citation is used to establish the new bounded-partial-inverse or componentwise-accuracy claims. The residual identity r=\alpha P h^2 used in Algorithm 5.1 is derived algebraically inside the paper from the Newton update, not imported as an assumption. Similarly, the quantity \||R^{-1}_{x^*}|x^*|\| in Theorem 6.2 comes from a floating-point error analysis of the residual computation and is compared with observed Newton stagnation levels rather than calibrated against them. A possible proof concern is that Theorem 4.4 invokes the boundedness property of S from Theorem 3.4, whose proof assumes irreducibility and constant off-diagonal entries; this is a hypothesis-verification or correctness issue, not a circularity, because no equation is identified with its own input by construction. No fitted parameter is renamed as a prediction, and no load-bearing claim is reduced to a self-citation. Hence no significant circularity is present.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper's theoretical results derive from known machinery: the matrix tree theorem, M-matrix theory, GTH, and fixed-point arguments. The only quantities that could be mistaken for free parameters (omega, kappa, ||abs(R^{-1}*x*)*x*||) are defined from the exact solution and the operator, not fitted to data. The main unstated input is the irreducibility condition ensuring the stability of the S component; the other assumptions (m>0, zero-sum perturbations) are stated.

assumptions (4)
  • standard math The all-minors matrix tree theorem (Theorem 3.1) and the GTH algorithm results of [2,15] hold as cited.
    Used to represent det M and adj M as sums of nonnegative tree products (Theorem 3.2 and 3.4) and to solve M-matrices without subtraction in Algorithms 5.1 and 5.2.
  • domain assumption The minimal solution satisfies m>0 in the problems considered (Section 4, after Theorem 2.1).
    Componentwise errors divide by m_i; positivity also keeps R_x a nonsingular M-matrix along the Newton path and avoids degenerate zero entries during iteration.
  • domain assumption Perturbations in Theorem 4.4 preserve the stochastic constraints 1^T a = 1-alpha and 1^T B_(1) = alpha * 1^T (Section 4.3).
    The zero-sum structure is what allows replacing R^{-1} with the bounded S, so the alpha=1/2 singularity disappears from the bound.
  • domain assumption The support graph of R_m^T is irreducible, or equivalently the quantities D_i in Theorem 3.4's proof are bounded away from zero along the alpha to 1/2 sequence (unstated in Theorem 4.4).
    Needed for S to remain bounded and omega in (4.8) to stay O(1); without it the headline 'ignore the singularity' claim can fail even though Theorem 4.4's bound is formally correct.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Accuracy and componentwise accuracy in multilinear PageRank." pith.science (2026). https://pith.science/paper/VMEE4JNU

@misc{pith2026250618356,
  author       = {Pith},
  title        = {Pith review of: Accuracy and componentwise accuracy in multilinear PageRank},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VMEE4JNU}},
  note         = {Machine review of arXiv:2506.18356}
}
abstract

We study the stability with respect to perturbations and the accuracy of numerical algorithms for computing solutions to the multilinear PageRank problem $\mathbf{x} = (1-\alpha)\mathbf{v} + \alpha \mathcal{P} \mathbf{x}^2$. Our results reveal that the solution can be more stable with respect to perturbations and numerical errors with respect to the classical bounds for nonlinear systems of equations (based on the norm of the Jacobian). In detail, one can obtain bounds for the minimal solution which ignore the singularity of the problem for $\alpha=1/2$, and one can show that the limiting accuracy of the Newton method depends not on the norm of the Jacobian but on a quantity that can be much smaller thanks to the nonnegativity structure of the equation. For the minimal solution, we also suggest subtraction-free modifications to the existing algorithms to achieve componentwise stability. Some of the theoretical results we obtain are interesting even outside the scope of this problem: bounds for more general quadratic vector equations, and a partial inverse for M-matrices which remains bounded when the matrix to invert approaches singularity.

Figures

Figures reproduced from arXiv: 2506.18356 by the authors.

Figure 1
Figure 1. Example 3: Results obtained from the computation of the minimal solution m. 0 10 20 30 40 50 60 70 80 90 100 10−16 10−14 10−12 10−10 10−8 10−6 10−4 10−2 100 Iterations N, enorm N, ecw NG, enorm NG, ecw BJ, enorm BJ, ecw BJGv, enorm BJGv, ecw NG, ∥rk∥∞ [PITH_FULL_IMAGE:figures/full_fig_p025_1.png] view at source ↗
Figure 2
Figure 2. Example 3: Results for the minimal solution m with α = 0.499999999999999. The residual for the NG method is also shown here (not invariant) [PITH_FULL_IMAGE:figures/full_fig_p025_2.png] view at source ↗
Figure 3
Figure 3. Example 4: Results obtained from the computation of the minimal solution m. 0.7 0.8 0.9 0.92 0.95 0.97 0.99 0.993 0.995 0.9951 0.99515 0.99517 0.995174 0.9951747 0.99517473 0.995174734 0.9951747342 0.99517473429 0.995174734291 0.9951747342915 0.99517473429157 0.995174734291573 0.9951747342915737 10−2 100 102 104 106 108 1010 1012 α N, enorm/ϵ N, ecw/ϵ NG, enorm/ϵ NG, ecw/ϵ cond(Rs ∗ ) ∥|R −1 s ∗ | s ∗∥ ∥R −1 s ∗ ∥∥s… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Example 5: Results for the computation of the stochastic solution s [PITH_FULL_IMAGE:figures/full_fig_p027_4.png]
Figure 5
Figure 5. Figure 5: Example 6: Results for the minimal solution m with α = 0.499999999999999. For this α, Example 4 yields cond(Rm∗ ) = 1.5382 · 1020, while Example 5 yields cond(Rm∗ ) = 1.8159 · 1020 [PITH_FULL_IMAGE:figures/full_fig_p028_5.png]
Figure 6
Figure 6. Figure 6: Example 6: Results for the minimal solution m with ε = 10−8 . 10−14 10−13 10−12 10−11 10−10 10−9 10−8 10−7 10−6 10−5 10−10 10−6 10−2 102 106 1010 ε ecw, upper bound, α = 0.9951747 ecw, upper bound, α = 0.99 ecw, upper bound, α = 0.8 [PITH_FULL_IMAGE:figures/full_fig_p…
Figure 7
Figure 7. Figure 7: Example 6: Results for the stochastic solution s, where P and v are taken from Example 5. It is interesting to note that the theoretical perturbation bound that we obtained are quite loose, and that in practice they seem to hold also for stochastic solutions, even thou…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. The mmatrix toolbox: componentwise accurate algorithms for M-matrices with triplet representation

    cs.MS 2026-07 accept novelty 5.0 of 10

    A Matlab/Fortran toolbox implements unblocked, recursive and blocked GTH algorithms that deliver componentwise-accurate linear algebra on M-matrices supplied via left or right triplets.

Reference graph

Works this paper leans on

22 extracted references · 17 canonical work pages · cited by 1 Pith paper

  1. [1]

    Abdesselam , The Grassmann–Berezin calculus and theorems of the matrix-tree type , Adv

    A. Abdesselam , The Grassmann–Berezin calculus and theorems of the matrix-tree type , Adv. Appl. Math., 33 (2004), pp. 51–70, https://doi.org/10.1016/j.aam.2003.07.002

    work page doi:10.1016/j.aam.2003.07.002 2004

  2. [2]

    A. S. Alfa, J. Xue, and Q. Ye , Entrywise perturbation theory for diagonally dominant M-matrices with applications, Numer. Math., 90 (2002), pp. 401–414, https://doi.org/10.1007/s002110100289

    work page doi:10.1007/s002110100289 2002

  3. [3]

    Batagelj and A

    V. Batagelj and A. Mrvar , Pajek datasets, 2006, http://vlado.fmf.uni-lj.si/pub/networks/data

    work page 2006

  4. [4]

    A. R. Benson, D. F. Gleich, and J. Leskovec , Tensor spectral clustering for partitioning higher- order network structures, in Proceedings of the 2015 SIAM International Conference on Data Mining, SIAM, 2015, pp. 118–126, https://doi.org/0.1137/1.9781611974010.14

    work page 2015

  5. [5]

    A. R. Benson, D. F. Gleich, and L.-H. Lim , The spacey random walk: a stochastic process for higher- order data, SIAM Rev., 59 (2017), pp. 321–345, https://doi.org/10.1137/16M1074023

    doi:10.1137/16m1074023 2017

  6. [6]

    Chaiken , A combinatorial proof of the all minors matrix tree theorem , SIAM J

    S. Chaiken , A combinatorial proof of the all minors matrix tree theorem , SIAM J. Alg. Disc. Meth., 3 (1982), pp. 319–329, https://doi.org/10.1137/0603033

    doi:10.1137/0603033 1982

  7. [7]

    Cipolla, M

    S. Cipolla, M. Redivo-Zaglia, and F. Tudisco , Extrapolation methods for fixed-point multilinear PageRank computations, Numer. Linear Algebra Appl., 27 (2020), p. e2280, https://doi.org/10.1002/ nla.2280

    work page 2020

  8. [8]

    T. A. Davis and Y. Hu , The university of Florida sparse matrix collection , ACM Transactions on Mathematical Software (TOMS), 38 (2011), pp. 1–25, https://doi.org/10.1145/2049662.2049663

    arXiv 2011

Show all 22 references
  1. [9]

    D. F. Gleich, L.-H. Lim, and Y. Yu , Multilinear PageRank, SIAM J. Matrix Anal. Appl., 36 (2015), pp. 1507–1541, https://doi.org/10.1137/140985160

    2015 doi

  2. [10]

    Grindrod and T

    P. Grindrod and T. Lee , Comparison of social structures within cities of very different sizes , Royal Society Open Science, 3 (2016), p. 150526, https://doi.org/10.1098/rsos.150526

    2016 doi

  3. [11]

    N. J. Higham , Accuracy and stability of numerical algorithms. , Philadelphia, PA: SIAM, 2nd ed. ed., 2002, https://doi.org/10.1137/1.9780898718027

    2002 doi

  4. [12]

    F. Lai, W. Li, X. Peng, and Y. Chen , Anderson accelerated fixed-point iteration for multilinear Page- Rank, Numer. Linear Algebra Appl., 30 (2023), p. e2499, https://doi.org/10.1002/nla.2499

    2023 doi

  5. [13]

    Meini and F

    B. Meini and F. Poloni , Perron-based algorithms for the multilinear PageRank, Numer. Linear Algebra Appl., 25 (2018), p. e2177, https://doi.org/10.1002/nla.2177

    2018 doi

  6. [14]

    C. B. Moler , Numerical computing with Matlab, SIAM, 2004

    2004

  7. [15]

    C. A. O’Cinneide , Relative-error bounds for the LU decomposition via the GTH algorithm , Numer. Math., 73 (1996), pp. 507–519, https://doi.org/10.1007/s002110050203

    1996 doi

  8. [16]

    Poloni , Quadratic vector equations , Linear Algebra Appl., 438 (2013), pp

    F. Poloni , Quadratic vector equations , Linear Algebra Appl., 438 (2013), pp. 1627–1644, https://doi. org/10.1016/j.laa.2011.05.036

    2013 doi

  9. [17]

    Saad , Iterative Methods for Sparse Linear Systems , SIAM, 2003

    Y. Saad , Iterative Methods for Sparse Linear Systems , SIAM, 2003

    2003

  10. [18]

    G. W. Stewart and J.-g. Sun , Matrix perturbation theory, Boston etc.: Academic Press, Inc., 1990

    1990

  11. [19]

    Stoer, R

    J. Stoer, R. Bulirsch, R. Bartels, W. Gautschi, and C. Witzgall , Introduction to Numerical Analysis, vol. 1993, Springer, 1980

    1993

  12. [20]

    Tisseur , Newton’s method in floating point arithmetic and iterative refinement of generalized eigen- value problems , SIAM J

    F. Tisseur , Newton’s method in floating point arithmetic and iterative refinement of generalized eigen- value problems , SIAM J. Matrix Anal. Appl., 22 (2001), pp. 1038–1057, https://doi.org/10.1137/ S0895479899359837, eprints.maths.manchester.ac.uk/979/1/tisseur3.pdf. ACCURA...

    2001

  13. [21]

    J. Xue, S. Xu, and R.-C. Li , Accurate solutions of M-matrix algebraic Riccati equations, Numer. Math., 120 (2012), pp. 671–700, https://doi.org/10.1007/s00211-011-0421-0

    2012 doi

  14. [22]

    J. Xue, S. Xu, and R.-C. Li , Accurate solutions of M-matrix Sylvester equations , Numer. Math., 120 (2012), pp. 639–670, https://doi.org/10.1007/s00211-011-0420-1

    2012 doi

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.