Pith. sign in

REVIEW 2 major objections 5 minor 37 references

Optimal structured approximation of Fourier subspaces, Toeplitz matrices, and exponential sums

T0 review · 2 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read One spectral-estimation algorithm, Gradient-MUSIC, solves Toeplitz matrix recovery, Fourier subspace estimation, and exponential-sum recovery at provably optimal rates, via a quantitative transference principle.

desk verdict Genuinely useful transference principle, but Theorem 3.1's exact-rank guarantee is not actually proved and the r-dependence in the optimality claims is loose. read the letter →

arxiv 2511.17239 v2 pith:PWYFMEHF submitted 2025-11-21 cs.IT math.IT

Albert Fannjiang , Weilin Li This is my paper · ORCID
classification cs.ITmath.IT MSC 15A2915B0565F55
keywords Gradient-MUSICToeplitzmatrixapproximationFouriersubspaceestimationspectralexponentialsumsminimaxoptimalityHankelmatricessuper-resolution
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 argues that three problems usually treated separately—recovering the frequencies of an exponential sum, approximating a noisy low-rank Toeplitz matrix, and recovering a Fourier subspace from a single snapshot—are the same problem in disguise. The authors show that one spectral-estimation algorithm, Gradient-MUSIC, can be used as a black box for all three, and that the resulting estimators match the information-theoretic lower bounds. For Toeplitz matrices the guarantee is ||T − T̂||₂ ≤ C√r||E||₂ under noise ||E||₂ ≤ αn, with no dependence on n except through the noise; for Fourier subspaces the guarantee is ||sin(U, Û)||_F ≤ C√(r/n)||z||₂. A sympathetic reader would care because these rates are provably the best any method can achieve, and the algorithms are computationally light—essentially the cost of one truncated SVD plus a gradient search.

What carries the argument

The load-bearing object is a Fourier matrix Φ(n,x) with rows indexed by integer times 0,...,n−1 and columns e^{ijx_k}; a separated frequency set x with Δ(x) ≥ 8π/n (or 16π/n) makes Φ well-conditioned via the large-sieve bound (2.2). Gradient-MUSIC is the subroutine: it forms a MUSIC-type objective on an input subspace and locates local minima by gradient descent, with the guarantee that a subspace perturbation of size δ leads to a frequency error at most 10δ/n. The transference principle is the chain of inequalities that turns frequency error into matrix or subspace error, and the lower-bound proofs that turn any too-good structured approximator into a too-good spectral estimator.

What would settle it

Construct a rank-2 Toeplitz matrix whose underlying frequencies are separated by less than 8π/n, add noise with ||E||₂ = αn for the α used in Theorem 3.1, and measure whether the output error exceeds C√r||E||₂; if it does, the conditioning premise fails. Alternatively, exhibit a method that on the full parameter class T(n,r) achieves an error strictly below the lower bound of Theorem 3.2, which would disprove the minimax claim.

Watch

Extended reading notes

Core claim

The central claim is that the structured-approximation problems reduce, through a quantitative transference principle, to spectral estimation, and that Gradient-MUSIC—an algorithm that takes an arbitrary subspace close to a Fourier subspace and returns the underlying frequencies—carries the optimal error rates across all three. Theorem 3.1 states that for a rank-r Toeplitz matrix T with a well-conditioned Fourier factorization, corrupted by arbitrary noise E with ||E||₂ ≤ αn, the estimator outputs a Toeplitz matrix of rank exactly r with spectral-norm error at most C√r||E||₂ and Frobenius error at most Cr||E||₂; Corollary 3.3 proves a matching minimax lower bound, so the rate is optimal in n

Load-bearing premise

The entire framework depends on the frequencies being sufficiently separated (at least 8π/n or 16π/n) so that the Fourier matrix is well-conditioned; if frequencies cluster, the conditioning bound fails and the error guarantees collapse.

Editorial extensions

If this is right

  • Low-rank Toeplitz approximation is solved at minimax-optimal accuracy: for the parameter class T(n,r), no method can beat the C√r||E||₂ rate in n and noise.
  • Fourier subspace recovery from a single noisy snapshot has the same optimality: error at most C√(r/n)||z||₂, with a matching minimax lower bound.
  • The returned Toeplitz matrix always has rank exactly r, unlike alternating-projection or sparse-Fourier methods that can overestimate the rank.
  • A quantitative transference principle converts guarantees for spectral estimation into guarantees for the other two problems, so all three share the same noise-versus-dimension trade-offs.
  • The estimator runs in O(n²r + nr² log(n/||E||₂)) time, roughly one truncated SVD plus a gradient-descent search.

Reading between the lines

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

  • Beyond the stated results, the transference principle suggests that any spectral-estimation method with a subspace-input guarantee and a matching lower bound would also yield optimal Toeplitz and Fourier-subspace estimators; Gradient-MUSIC is one instantiation, not the only possible one.
  • The authors leave open whether the extra 1/√r factor in the noise assumption for Fourier subspace estimation is necessary; removing it would extend the guarantee to higher-rank settings.
  • Because the truncated SVD already achieves minimax optimality as a preprocessing step, one testable extension is to plug the same pipeline into other structured families whose leading eigenspaces are parameterized by a small set of continuous parameters.
  • A natural next problem is the converse of the transference: whether an optimal Toeplitz or subspace estimator can be converted back into an optimal spectral estimator, which the paper does not answer.
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.

Referee Report

2 major / 5 minor

Summary. The paper studies two structured approximation problems: recovering a low-rank Toeplitz matrix T from a noisy observation T+E, and recovering the Fourier subspace range(Φ(n,x)) from a single noisy vector Φ(n,x)a+z. The proposed algorithms first compute a leading singular subspace, feed it to the authors' Gradient-MUSIC spectral-estimation method, and then form a Toeplitz matrix or a Fourier subspace from the estimated frequencies. The main theorems (3.1 and 4.2) give error bounds of order √r∥E∥₂ and √(r/n)∥z∥₂ under separation and small-noise assumptions, and Theorems 3.2 and 4.3 give lower bounds showing optimality in n and in the noise level. The paper also contains complexity estimates and numerical experiments comparing the Toeplitz estimator with alternating projection.

Significance. The transference principle — using a spectral-estimation algorithm as a black box for structured matrix and subspace recovery — is conceptually attractive and, if the technical gaps are fixed, would provide the first provably minimax-optimal polynomial-time methods for these two problems. The algorithms are deterministic, have explicit complexity bounds, and the authors provide reproducible code. The reliance on the authors' prior Gradient-MUSIC analysis [12] is heavy, but that prior work is stated to contain parameter-free, assumption-explicit proofs. The main risk is the exact-rank guarantee in Theorem 3.1, which is not established by the given argument.

major comments (2)
  1. [Sec. 3.2, Theorem 3.1] The 'rank exactly r' guarantee is not supported. The estimator is T̂ = Φ̂ diag(â) Φ̂*, so its rank equals the number of nonzero entries of â (Φ̂ has full column rank). The proof obtains ∥â−a∥∞ ≲ √r∥E∥₂/n (Eq. (3.10)), which under ∥E∥₂≤αn is ≤ Cα√r. Since |a_j|≥1, one can conclude â_j≠0 only if Cα√r<1. But α is declared absolute while r is allowed to be arbitrarily large (r<n), so for r > (Cα)^{-2} the argument fails; an adversarial E satisfying the norm bound can in principle zero out an amplitude. The sentence 'This is clearly a Toeplitz matrix of rank exactly r' is therefore unjustified. This gap propagates to Corollary 3.3 and the abstract's exact-rank claim. The theorem must be restated with an explicit r-dependent smallness condition, or the proof must provide a different mechanism for preserving the rank.
  2. [Sec. 3.3, proof of Theorem 3.2] The reduction from spectral estimation to Toeplitz recovery is not fully specified. The observation is written as y=Φ(2n−1,x)a+z with the paper's convention that Fourier matrices have rows j=0,...,2n−2, but the Toeplitz matrix T(y) needs entries t_m for m=−(n−1),...,n−1. It is not explained how the negative-index entries are obtained from y. If a symmetric index set is intended, this must be stated explicitly and the spectral-estimation lower bound (Lemma 2.2) must be applied to that convention. As written, the equality T(y)=Φ(n,x)diag(a)Φ(n,x)^*+T(z) is not demonstrated.
minor comments (5)
  1. [Abstract / title] The first abstract says 'three structured approximation problems' and gives the Toeplitz bound as C∥E∥₂ without √r, while the full-text abstract and Theorem 3.1 state two problems and give C√r∥E∥₂. Please align these statements.
  2. [Sec. 4, Theorem 4.2] The statement says the estimator outputs a Fourier subspace bU ⊆ C^n of dimension n; it should be dimension r.
  3. [Sec. 4.3, proof of Theorem 4.3 and Corollary 4.4] The line 'Since C = 100' should read 'Since C = 1/100'; with C=1/100 the displayed inequality contradicts Lemma 2.2. Also, Corollary 4.4 uses F*(m,r,...) in the proof although the statement uses n.
  4. [Sec. 4.1] The paper motivates the noise condition ∥z∥₂≤α√(n/r) by noting that ∥z∥₂≤α√n is necessary, but does not know if the extra 1/√r is needed. This is fine, but the gap between the necessary and sufficient conditions should be highlighted in the discussion.
  5. [Sec. 3.2] The claim that r can be deduced by a thresholding procedure cites [12, Lemma 5.6] but the lemma is not stated. Since r is assumed known in the algorithm and theorems, the reader would benefit from a short statement of the thresholding guarantee.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: the Toeplitz/subspace results are genuine reductions to prior parameter-free spectral-estimation bounds, not restatements of their inputs.

full rationale

The paper's new upper bounds (Theorems 3.1 and 4.2) are built from explicit estimators, bT = bPhi diag(ba)bPhi* or bU = range(Phi(n,bx)). The error is controlled by triangle inequalities together with the frequency bound (3.8) derived from Theorem 2.1 and the amplitude bound (3.10) from [12, Lemma 5.12]. These cited results are parameter-free: Theorem 2.1 assumes only ||sin(U,W)||_2 <= 1/100 for an arbitrary subspace W, and Lemma 2.2 is a minimax bound over general spectral-estimation data; neither is tailored to Toeplitz matrices or includes the target approximation problem as an input. Thus the borrowing is a normal, load-bearing citation rather than a circular reduction. The lower bounds (Theorems 3.2 and 4.3) are embeddings from the spectral-estimation minimax problem; the identity T(y) = Phi(n,x)diag(a)Phi(n,x)* + T(z) with ||T(z)||_2 <= alpha n is exact, and the contradiction argument uses Lemma 2.2 as an external lower bound, not as the conclusion being proved. The defining class T(n,r) is restricted to (x,a) in P(8pi/n,r); this is an explicit regularity model, not a way of hiding the target inside the input. No fitted parameter is subsequently renamed as a prediction, and no uniqueness/ansatz result is imported from the authors to force a choice. The only reservations are non-circular correctness/scope issues: (i) the 'rank exactly r' guarantee in Theorem 3.1 is not established as stated, because the proof gives ||ba-a||_inf <= sqrt(r)||E||_2/n <= alpha sqrt(r); for fixed absolute alpha and large r this can exceed min_j |a_j| >= 1, so the construction bPhi diag(ba) bPhi* can lose rank, contradicting the paper's assertion 'This is clearly a Toeplitz matrix of rank exactly r' in Section 3.2; (ii) the supplied abstract claims optimality in n, r, and ||E||_2, while the full-text abstract and Corollary 3.3 only show optimality in n and ||E||_2 with a sqrt(r) gap between lower bound alpha n/320 and upper bound C sqrt(r) alpha n; (iii) the minimum-separation restriction is explicitly acknowledged as a limitation. These are correctness or scope weaknesses, not circularity.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new entities or physical postulates. Its free parameters are mostly problem-class restrictions (amplitude range, separation threshold, noise-to-ratio α, rank). The '10' is arbitrary; α and the separation thresholds are genuine restrictions.

free parameters (4)
  • amplitude dynamic-range bound (10 in P(h,r)) = 10
    The parameter set P(h,r) restricts amplitudes to 1≤|a_j|≤10. The upper bound '10' is chosen for convenience and only affects absolute constants, but it is a restriction on the problem class.
  • minimum separation threshold β (8π/n or 16π/n) = 8π/n (Toeplitz), 16π/n (subspace)
    The condition Δ(x) ≥ 8π/n or 16π/n is required to use the conditioning inequality (2.2). The specific constant 8 is arbitrary and affects all constants in the proofs.
  • noise-to-signal constant α = unspecified, 'sufficiently small'
    The theorems require ‖E‖₂ ≤ αn (or ‖z‖₂ ≤ α√(n/r)) for a sufficiently small absolute α. The paper never specifies the numerical value of α, and all the main bounds hold only for this implicit range.
  • r (rank) assumed known = unknown
    The algorithms require the true rank r as input. The paper says r can be determined by a thresholding procedure from prior work ([12, Lemma 5.6]), but this is not analyzed here and is effectively an assumption for the main theorems.
assumptions (6)
  • standard math The Vandermonde determinant theorem for the identifiability lemma (Lemma 4.1)
    Used to prove uniqueness of the frequency set x when n ≥ 2r.
  • standard math Wedin's sine-theta theorem (perturbation theory)
    Used throughout Theorems 3.1, 4.2, and the lower-bound proofs to convert matrix perturbation to subspace perturbation.
  • standard math Weyl's inequality for singular values
    Used in Theorem 3.2 to show the gap between singular values of T+E.
  • domain assumption Conditioning of Vandermonde/Fourier matrices with well-separated nodes (Aubel-Bölcskei [2])
    The bound √(n(1-β^{-1})) ≤ σ_min(Φ) ≤ σ_max(Φ) ≤ √(n(1+β^{-1})) for Δ(x) ≥ 2πβ/n is external and controls all error propagation.
  • domain assumption The full correctness of Gradient-MUSIC (Theorem 2.1, from the authors' prior work [12])
    The whole paper treats Gradient-MUSIC as a black-box oracle. Any flaw in [12] would invalidate Theorems 3.1 and 4.2.
  • domain assumption The thresholding procedure for determining r ([12, Lemma 5.6]) applicable to the Hankel/Toeplitz data
    The paper says the rank can be deduced via thresholding, but the analysis is not reproduced. The main theorems assume r is known.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Optimal structured approximation of Fourier subspaces, Toeplitz matrices, and exponential sums." pith.science (2026). https://pith.science/paper/PWYFMEHF

@misc{pith2026251117239,
  author       = {Pith},
  title        = {Pith review of: Optimal structured approximation of Fourier subspaces, Toeplitz matrices, and exponential sums},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PWYFMEHF}},
  note         = {Machine review of arXiv:2511.17239}
}
abstract

This paper studies three structured approximation problems: (1) Recovering the range of a Fourier matrix from a single observation, (2) Recovering a corrupted low-rank Toeplitz/Hankel matrix, and (3) Recovering a finite exponential sum from noisy samples. All three problems are computationally challenging because their structural constraints are difficult to enforce directly. We show that all three tasks can be solved efficiently and optimally by applying the Gradient-MUSIC algorithm for spectral estimation. To provide an example, for a rank-$r$ Toeplitz matrix $T\in {\mathbb C}^{n\times n}$ that satisfies a regularity assumption and is corrupted by an arbitrary $E\in {\mathbb C}^{n\times n}$ such that $\|E\|_2\leq \alpha n$, our algorithm outputs a Toeplitz matrix $T_\sharp$ of rank exactly $r$ such that $\|T-T_\sharp\|_2 \leq C \|E\|_2$, where $C,\alpha>0$ are absolute constants. This performance guarantee is minimax optimal in $n$, $r$, and $\|E\|_2$. For the other two structured approximation problems, we also provide algorithms that are minimax optimal in the number of samples, rank/sparsity, and noise level. At the heart of this paper is a quantitative transference principle which shows how to convert computational methods and theory for spectral estimation into corresponding methods and theory for the other three problems.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 1 linked inside Pith

  1. [12]

    Optimality of Gradient-MUSIC for spectral estimation

    Albert Fannjiang, Weilin Li, and Wenjing Liao. Optimality of Gradient-MUSIC for spectral estimation. arXiv preprint arXiv:2504.06842 , 2025

    arXiv 2025

  2. [1]

    Alternating projections on nontangential manifolds

    Fredrik Andersson and Marcus Carlsson. Alternating projections on nontangential manifolds. Constructive Approximation, 38(3):489–525, 2013

    2013

  3. [2]

    Vandermonde matrices with nodes in the unit disk and the large sieve

    C´ eline Aubel and Helmut B¨ olcskei. Vandermonde matrices with nodes in the unit disk and the large sieve. Applied and Computational Harmonic Analysis , 47(1):53–86, 2019

    2019

  4. [3]

    Alex H. Barnett. How exponentially ill-conditioned are contiguous submatrices of the Fourier matrix? SIAM Review, 64(1):105–131, 2022

    2022

  5. [4]

    Conditioning of partial nonuniform Fourier matrices with clustered nodes

    Dmitry Batenkov, Laurent Demanet, Gil Goldman, and Yosef Yomdin. Conditioning of partial nonuniform Fourier matrices with clustered nodes. SIAM Journal on Matrix Analysis and Applications, 41(1):199–220, 2020

    2020

  6. [5]

    The spectral prop- erties of Vandermonde matrices with clustered nodes

    Dmitry Batenkov, Benedikt Diederichs, Gil Goldman, and Yosef Yomdin. The spectral prop- erties of Vandermonde matrices with clustered nodes. Linear Algebra and its Applications , 609:37–72, 2021. 19

    2021

  7. [6]

    James A. Cadzow. Signal enhancement-a composite property mapping algorithm. IEEE Trans- actions on Acoustics, Speech, and Signal Processing , 36(1):49–62, 1988

    1988

  8. [7]

    Spectral compressed sensing via structured matrix completion

    Yuxin Chen and Yuejie Chi. Spectral compressed sensing via structured matrix completion. In International Conference on Machine Learning , pages 414–422. PMLR, 2013

    2013

Show all 37 references
  1. [8]

    Spectral methods for data science: A statistical perspective

    Yuxin Chen, Yuejie Chi, Jianqing Fan, Cong Ma, et al. Spectral methods for data science: A statistical perspective. Foundations and Trends® in Machine Learning , 14(5):566–806, 2021

    2021

  2. [9]

    Chu, Robert E Funderlic, and Robert J

    Moody T. Chu, Robert E Funderlic, and Robert J. Plemmons. Structured low rank approxi- mation. Linear Algebra and its Applications , 366:157–172, 2003

    2003

  3. [10]

    Moment problems and low rank Toeplitz approximations

    George Cybenko. Moment problems and low rank Toeplitz approximations. Circuits, Systems and Signal Processing, 1(3):345–366, 1982

    1982

  4. [11]

    Ioffe, and Adrian S

    Dmitriy Drusvyatskiy, Alexander D. Ioffe, and Adrian S. Lewis. Transversality and alternating projections for nonconvex sets. Foundations of Computational Mathematics , 15(6):1637–1651, 2015

    2015

  5. [13]

    Hankel low-rank approximation and completion in time series analysis and forecasting: a brief review

    Jonathan Gillard and Konstantin Usevich. Hankel low-rank approximation and completion in time series analysis and forecasting: a brief review. Statistics and Its Interface , 16(2):287–303, 2023

    2023

  6. [14]

    Toeplitz Forms and Their Applications

    Ulf Grenander and Gabor Sz¨ ego. Toeplitz Forms and Their Applications . University of Cali- fornia Press, 1958

    1958

  7. [15]

    Grigoriadis, Arthur E

    Karolos M. Grigoriadis, Arthur E. Frazho, and Robert E. Skelton. Application of alternating convex projection methods for computation of positive Toeplitz matrices. IEEE Transactions on Signal Processing, 42(7):1873–1875, 2002

    2002

  8. [16]

    Combinatorial Theory

    Marshall Hall. Combinatorial Theory. John Wiley & Sons, 1967

    1967

  9. [17]

    Deconvolution and regularization with Toeplitz matrices

    Per Christian Hansen. Deconvolution and regularization with Toeplitz matrices. Numerical Algorithms, 29(4):323–378, 2002

    2002

  10. [18]

    Toeplitz low-rank approximation with sublinear query complexity

    Michael Kapralov, Hannah Lawrence, Mikhail Makarov, Cameron Musco, and Kshiteej Sheth. Toeplitz low-rank approximation with sublinear query complexity. In Proceedings of the 2023 Annual ACM-SIAM Symposium on Discrete Algorithms (SODA) , pages 4127–4158. SIAM, 2023

    2023

  11. [19]

    Two decades of array signal processing research: the parametric approach

    Hamid Krim and Mats Viberg. Two decades of array signal processing research: the parametric approach. IEEE Signal Processing Magazine , 13(4):67–94, 1996

    1996

  12. [20]

    On the smallest singular value of multivariate Vandermonde matrices with clustered nodes

    Stefan Kunis and Dominik Nagel. On the smallest singular value of multivariate Vandermonde matrices with clustered nodes. Linear Algebra and its Applications , 604:1–20, 2020

    2020

  13. [21]

    Toeplitz structured subspace for multi-channel blind identification methods

    Abdulmajid Lawal, Qadri Mayyala, Karim Abed-Meraim, Naveed Iqbal, and Azzedine Zer- guine. Toeplitz structured subspace for multi-channel blind identification methods. Signal Processing, 188:108152, 2021. 20

    2021

  14. [22]

    Multiscale estimates for the condition number of non-harmonic fourier matrices

    Weilin Li. Multiscale estimates for the condition number of non-harmonic fourier matrices. Mathematics of Computation , 94(356):2895–2929, 2025

    2025

  15. [23]

    Stable super-resolution limit and smallest singular value of restricted Fourier matrices

    Weilin Li and Wenjing Liao. Stable super-resolution limit and smallest singular value of restricted Fourier matrices. Applied and Computational Harmonic Analysis , 51:118–156, 2021

    2021

  16. [24]

    Sublinear time low-rank approximation of Toeplitz ma- trices

    Cameron Musco and Kshiteej Sheth. Sublinear time low-rank approximation of Toeplitz ma- trices. In Proceedings of the 2024 Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), pages 5084–5117. SIAM, 2024

    2024

  17. [25]

    A fast algorithm for convolutional structured low-rank matrix recovery

    Gregory Ongie and Mathews Jacob. A fast algorithm for convolutional structured low-rank matrix recovery. IEEE Transactions on Computational Imaging , 3(4):535–550, 2017

    2017

  18. [26]

    Identification of nonlinear systems using polynomial nonlinear state space models

    Johan Paduart, Lieve Lauwers, Jan Swevers, Kris Smolders, Johan Schoukens, and Rik Pin- telon. Identification of nonlinear systems using polynomial nonlinear state space models. Automatica, 46(4):647–656, 2010

    2010

  19. [27]

    Ben Rosen

    Haesun Park, Lei Zhang, and J. Ben Rosen. Low rank approximation of a Hankel matrix by structured total least norm. BIT Numerical Mathematics , 39(4):757–779, 1999

    1999

  20. [28]

    A robust sparse Fourier transform in the continuous setting

    Eric Price and Zhao Song. A robust sparse Fourier transform in the continuous setting. In 56th Annual IEEE Symposium on Foundations of Computer Science (FOCS) , pages 583–600. IEEE, 2015

    2015

  21. [29]

    Ben Rosen, Haesun Park, and John Glick

    J. Ben Rosen, Haesun Park, and John Glick. Total least norm formulation and solution for structured problems. SIAM Journal on Matrix Analysis and Applications, 17(1):110–126, 1996

    1996

  22. [30]

    Ralph O. Schmidt. Multiple emitter location and signal parameter estimation. IEEE Trans- actions on Antennas and Propagation , 34(3):276–280, 1986

    1986

  23. [31]

    T. J. Suffridge and Tom L. Hayden. Approximation by a Hermitian positive semidefinite Toeplitz matrix. SIAM Journal on Matrix Analysis and Applications , 14(3):721–734, 1993

    1993

  24. [32]

    Joel A. Tropp. User-friendly tail bounds for sums of random matrices. Foundations of Com- putational Mathematics , 12(4):389–434, 2012

    2012

  25. [33]

    High-dimensional Probability: An Introduction with Applications in Data Science, volume 47

    Roman Vershynin. High-dimensional Probability: An Introduction with Applications in Data Science, volume 47. Cambridge University Press, 2018

    2018

  26. [34]

    and Ji-Guang Sun

    Gilbert W. and Ji-Guang Sun. Matrix Perturbation Theory . Academic Press Boston, 1990

    1990

  27. [35]

    Seismic sparse-spike deconvolution via Toeplitz-sparse matrix factorization

    Lingling Wang, Qian Zhao, Jinghuai Gao, Zongben Xu, Michael Fehler, and Xiudi Jiang. Seismic sparse-spike deconvolution via Toeplitz-sparse matrix factorization. Geophysics, 81(3):V169–V182, 2016

    2016

  28. [36]

    Sparse inverse synthetic aperture radar imaging using structured low-rank method

    Gang Xu, Bangjie Zhang, Jianlai Chen, Fan Wu, Jialian Sheng, and Wei Hong. Sparse inverse synthetic aperture radar imaging using structured low-rank method. IEEE Transactions on Geoscience and Remote Sensing , 60:1–12, 2021

    2021

  29. [37]

    Vandermonde factorization of Hankel matrix for complex exponential signal recovery—Application in fast NMR spectroscopy

    Jiaxi Ying, Jian-Feng Cai, Di Guo, Gongguo Tang, Zhong Chen, and Xiaobo Qu. Vandermonde factorization of Hankel matrix for complex exponential signal recovery—Application in fast NMR spectroscopy. IEEE Transactions on Signal Processing , 66(21):5520–5533, 2018. 21

    2018

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.