Pith. sign in

REVIEW 2 major objections 2 minor 57 references

Low-Rank Tensor Completion Based on Fractional Regularization with Ky Fan p-k Norm

T0 review · 2 major / 2 minor · reviewed 2026-06-26 · grok-4.3

Pith's one-line read The TNPK ratio of tensor nuclear norm to Ky Fan p-k norm approximates tubal rank and makes low-rank tensors local minimizers under the null space property.

desk verdict New TNPK ratio surrogate for tubal rank with local-minimizer proof under tensor NSP and an ADMM solver, but the NSP is not shown to hold for the sampling operators in the experiments. read the letter →

arxiv 2606.19046 v1 pith:WYG7NJJ5 submitted 2026-06-17 cs.CV

classification cs.CV
keywords low-ranktensorcompletiontubalrankKyFanp-knormnonconvexsurrogatenuclearnullspacepropertyADMMalgorithmfractionalregularization
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

The paper introduces TNPK, a nonconvex surrogate formed as the ratio of the tensor nuclear norm to the tensor Ky Fan p-k norm, to approximate the tensor tubal rank for low-rank tensor completion. This surrogate is scale-invariant, admits closed-form solutions for certain parameter choices, and reduces to the TNK or TNF ratios under specific p and k. The authors build a completion model and prove that low-rank tensors are its local minimizers whenever the tensor null space property holds for the measurement operator. They derive the proximal operator of the Ky Fan p-k inverse-norm and give an ADMM algorithm with subsequential convergence. Experiments on synthetic and real-world data show improved recovery over competing methods.

What carries the argument

The TNPK surrogate (ratio of tensor nuclear norm to tensor Ky Fan p-k norm), which acts as a nonconvex, scale-invariant approximation to tensor tubal rank with closed-form solutions for chosen p and k.

What would settle it

A concrete measurement operator obeying the tensor NSP for which some low-rank tensor fails to be a local minimizer of the TNPK-regularized objective, or a dataset where the method shows no recovery advantage over convex baselines despite NSP holding.

Watch

Extended reading notes

Core claim

The TNPK surrogate accurately approximates the tensor tubal rank. Under the tensor null space property, low-rank tensors are local minimizers of the resulting LRTC model. The approach includes the proximal operator for the Ky Fan p-k inverse-norm and an ADMM solver with guaranteed subsequential convergence under mild conditions.

Load-bearing premise

The linear measurement operator in the tensor completion problem satisfies the tensor null space property.

Editorial extensions

If this is right

  • Low-rank tensors can be recovered more accurately than with convex nuclear-norm surrogates.
  • Choices of p and k recover the TNK and TNF ratios as special cases, giving modeling flexibility.
  • The ADMM algorithm converges subsequentially under mild conditions.
  • Empirical gains appear on both synthetic tensors and real-world data such as images.

Reading between the lines

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

  • The same ratio construction could be tested on related problems such as tensor robust principal component analysis.
  • Varying p and k systematically might reveal which settings best match particular data modalities or noise levels.
  • Analogous ratio surrogates might be worth exploring for matrix completion to check whether the local-minimizer guarantee transfers.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 2 minor

Summary. The paper proposes the TNPK surrogate (ratio of tensor nuclear norm to tensor Ky Fan p-k norm) as a nonconvex approximation to tensor tubal rank for low-rank tensor completion. It proves that, under the tensor null space property, low-rank tensors are local minimizers of the resulting LRTC model; derives the proximal operator of the Ky Fan p-k inverse-norm; develops an ADMM solver with subsequential convergence; and reports experimental superiority on synthetic and real datasets, with reductions to TNK and TNF for specific p and k.

Significance. If the claims hold, the work offers a scale-invariant, flexible nonconvex regularizer with explicit reductions to prior surrogates, conditional local-minimizer guarantees, and a convergent algorithm. The experimental gains would strengthen the case for fractional Ky-Fan-based surrogates in tensor recovery.

major comments (2)
  1. [Theoretical analysis / proof of local minimizers] Theoretical analysis section (proof under tensor NSP): The local-minimizer guarantee is conditional on the tensor null space property holding for the linear measurement operator. The manuscript provides no explicit incoherence conditions, sampling-rate bounds, or numerical verification that the entry-sampling operators used in the experiments satisfy NSP at the reported tensor dimensions and sampling ratios. This disconnects the theorem from the practical recovery instances where superiority is claimed.
  2. [TNPK definition and experimental setup] Model and parameter selection (Section on TNPK definition and experiments): p and k are free parameters whose choices affect both the surrogate and the proximal operator. The manuscript does not report a systematic selection procedure or sensitivity analysis; if these are tuned post-hoc on the test sets, the reported gains may not generalize.
minor comments (2)
  1. [Proximal operator derivation] Notation: the distinction between the tensor Ky Fan p-k norm and its inverse-norm should be stated explicitly when the proximal operator is introduced.
  2. [Experimental results] Experiments: error bars or multiple random seeds are not mentioned for the synthetic data results; adding them would strengthen the superiority claims.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive feedback. We address each major comment below and outline the revisions we will make.

read point-by-point responses
  1. Referee: Theoretical analysis section (proof under tensor NSP): The local-minimizer guarantee is conditional on the tensor null space property holding for the linear measurement operator. The manuscript provides no explicit incoherence conditions, sampling-rate bounds, or numerical verification that the entry-sampling operators used in the experiments satisfy NSP at the reported tensor dimensions and sampling ratios. This disconnects the theorem from the practical recovery instances where superiority is claimed.

    Authors: We agree that the local-minimizer result is conditional on the tensor NSP. Deriving explicit sampling-rate bounds or incoherence conditions for the nonconvex TNPK surrogate is technically involved and lies outside the primary scope of the paper, which focuses on introducing the surrogate, proving the conditional property, and developing the algorithm. In the revised manuscript we will add a clarifying remark in the theoretical section stating that the NSP is an assumption on the measurement operator and that empirical validation of NSP for the specific sampling operators used in experiments is left for future work. No numerical verification of NSP will be added, as computing it exactly is intractable for the tensor dimensions considered. revision: partial

  2. Referee: Model and parameter selection (Section on TNPK definition and experiments): p and k are free parameters whose choices affect both the surrogate and the proximal operator. The manuscript does not report a systematic selection procedure or sensitivity analysis; if these are tuned post-hoc on the test sets, the reported gains may not generalize.

    Authors: We acknowledge the need for clearer documentation of parameter selection. In the original experiments, p and k were chosen either to recover known special cases (p=1, k=1 yields TNF; p=1 yields TNK) or via a small grid search on a held-out validation subset of each dataset. We will revise the experimental section to explicitly describe this procedure, include a sensitivity analysis table showing performance variation across a range of p and k values, and move any test-set tuning details to supplementary material to ensure the reported gains are reproducible. revision: yes

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity detected; derivation is conditional on external NSP assumption

full rationale

The central result is a conditional proof: under the tensor null space property (an assumption imported from compressed sensing literature), low-rank tensors are local minimizers of the TNPK-based model. This does not reduce the claim to a self-definition, fitted parameter, or self-citation chain. The TNPK surrogate is explicitly constructed as a novel ratio, with reductions to TNK/TNF shown as special cases rather than hidden fits. The proximal operator and ADMM convergence are derived directly from the model without circular renaming or ansatz smuggling. No load-bearing self-citation or uniqueness theorem from the authors' prior work is invoked to force the result. The paper is self-contained against its stated assumptions.

Assumptions & free parameters 1 free parameters · 1 assumptions · 1 invented entities

The central claim rests on the tensor NSP as a domain assumption and introduces p and k as tunable parameters in the new TNPK surrogate; the surrogate itself is an invented entity without independent evidence outside the paper.

free parameters (1)
  • p and k
    Parameters controlling the Ky Fan p-k norm in the TNPK ratio; specific choices reduce to TNF or TNK.
assumptions (1)
  • domain assumption Tensor null space property (NSP) holds for the measurement operator
    Invoked to prove that low-rank tensors are local minimizers of the model.
invented entities (1)
  • TNPK surrogate (ratio of tensor nuclear norm to tensor Ky Fan p-k norm)
    purpose: Nonconvex approximation to tensor tubal rank with scale invariance and closed-form properties
    Newly proposed fractional regularization; no independent falsifiable evidence provided in abstract.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Low-Rank Tensor Completion Based on Fractional Regularization with Ky Fan p-k Norm." pith.science (2026). https://pith.science/paper/WYG7NJJ5

@misc{pith2026260619046,
  author       = {Pith},
  title        = {Pith review of: Low-Rank Tensor Completion Based on Fractional Regularization with Ky Fan p-k Norm},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WYG7NJJ5}},
  note         = {Machine review of arXiv:2606.19046}
}
read the original abstract

This paper addresses low-rank tensor completion (LRTC) by proposing a novel nonconvex surrogate, namely the ratio of the tensor nuclear norm to the tensor Ky Fan p-k norm (TNPK), to accurately approximate the tensor tubal rank. The TNPK possesses appealing properties, including scale invariance, parameter flexibility, and the existence of closed-form solutions under specific choices of p and k. With specific parameter settings of p and k, it reduces to the ratio of the tensor nuclear norm to the tensor Ky Fan k norm (TNK) or the ratio of the tensor nuclear norm to the tensor Frobenius norm (TNF). We construct a LRTC model and, under the tensor null space property (NSP), prove that low-rank tensors are local minimizers of the proposed model. Moreover, we derive the proximal operator of the Ky Fan p-k inverse-norm and further develop an efficient alternating direction method of multipliers (ADMM) algorithm with guaranteed subsequential convergence under mild conditions. Extensive experiments on synthetic and real-world datasets validate the superior performance of our method against state-of-the-art competitors.

Figures

Figures reproduced from arXiv: 2606.19046 by the authors.

Figure 1
Figure 1. Comparison of the approximation ability of different [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Comparison of results for different maximum numbers [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 4
Figure 4. presents the RSE, recovered tubal rank, and running time for different values of k under the three transforms. The results indicate that not all values of k yield satisfactory outcomes, and the final performance varies with k. Therefore, the choice of parameter k should be determined according to the specific application scenario. 4) Phase Transition Diagrams for Different Tubal Ranks and Sampling Rates: To verify t… view at source ↗
Figures from the paper (9 more)
Figure 3
Figure 3. Figure 3: Parameter Sensitivity Analysis of TNK (p=1, k=18) and [PITH_FULL_IMAGE:figures/full_fig_p013_3.png]
Figure 5
Figure 5. Figure 5: Comparison of success rates for the tensor completion [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: PSNR values of different models in color image [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: Comparison of evaluation metric values of various [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]
Figure 8
Figure 8. Figure 8: Comparison of recovery performance for four color images with randomly missing pixels. From left to right: observed [PITH_FULL_IMAGE:figures/full_fig_p016_8.png]
Figure 9
Figure 9. Figure 9: Comparison of recovery performance for six color images under grid and text mask corruption. From left to right: [PITH_FULL_IMAGE:figures/full_fig_p016_9.png]
Figure 10
Figure 10. Figure 10: Comparison of recovery performance for five multispectral images under 10% random sampling. From left to right: [PITH_FULL_IMAGE:figures/full_fig_p017_10.png]
Figure 11
Figure 11. Figure 11: Comparison of recovery performance for four videos under 10% random sampling. From left to right: observed image, [PITH_FULL_IMAGE:figures/full_fig_p018_11.png]
Figure 12
Figure 12. Figure 12: Convergence behavior of the proposed TNPK regu [PITH_FULL_IMAGE:figures/full_fig_p018_12.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

57 extracted references · 2 canonical work pages

  1. [1]

    Tensor robust kernel pca for multidimensional data,

    J. Lin, T.-Z. Huang, X.-L. Zhao, T.-Y . Ji, and Q. Zhao, “Tensor robust kernel pca for multidimensional data,” IEEE Trans. Neural Netw. Learn. Syst., vol. 36, no. 2, pp. 2662–2674, 2024

  2. [2]

    Nonnegative low rank tensor approximations with multidimensional image applications,

    T.-X. Jiang, M. K. Ng, J. Pan, and G.-J. Song, “Nonnegative low rank tensor approximations with multidimensional image applications,” Numer. Math., vol. 153, no. 1, pp. 141–170, 2023

  3. [3]

    Tensorial extensions of independent component analysis for multisubject fmri analysis,

    C. F. Beckmann and S. M. Smith, “Tensorial extensions of independent component analysis for multisubject fmri analysis,”NeuroImage, vol. 25, no. 1, pp. 294–311, 2005

  4. [4]

    Dual-enhanced high-order self-learning tensor singular value decomposition for robust principal component analysis,

    H. Xu, C. Fang, R. Wang, S. Chen, and J. Zheng, “Dual-enhanced high-order self-learning tensor singular value decomposition for robust principal component analysis,” IEEE Trans. Artif. Intell. , vol. 5, no. 7, pp. 3564–3578, 2024

  5. [5]

    Tensor learning meets dynamic anchor learning: From complete to incomplete multiview clustering,

    Y . Chen, X. Zhao, Z. Zhang, Y . Liu, J. Su, and Y . Zhou, “Tensor learning meets dynamic anchor learning: From complete to incomplete multiview clustering,” IEEE Trans. Neural Netw. Learn. Syst. , vol. 35, no. 11, pp. 15 332–15 345, 2023

  6. [6]

    Compressive spectral light field image reconstruction via online tensor representation,

    M. Marquez, H. Rueda-Chacon, and H. Arguello, “Compressive spectral light field image reconstruction via online tensor representation,” IEEE Trans. Image Process., vol. 29, pp. 3558–3568, 2020

  7. [7]

    Robust thick cloud removal for multitemporal remote sensing images using coupled tensor factorization,

    J. Lin, T.-Z. Huang, X.-L. Zhao, Y . Chen, Q. Zhang, and Q. Yuan, “Robust thick cloud removal for multitemporal remote sensing images using coupled tensor factorization,” IEEE Trans. Geosci. Remote Sens. , vol. 60, pp. 1–16, 2022

  8. [8]

    Tensor completion for estimating missing values in visual data,

    J. Liu, P. Musialski, P. Wonka, and J. Ye, “Tensor completion for estimating missing values in visual data,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 35, no. 1, pp. 208–220, 2012

Show all 57 references
  1. [9]

    Generalized nonconvex approach for low-tubal-rank tensor recovery,

    H. Wang, F. Zhang, J. Wang, T. Huang, J. Huang, and X. Liu, “Generalized nonconvex approach for low-tubal-rank tensor recovery,” IEEE Trans. Neural Netw. Learn. Syst. , vol. 33, no. 8, pp. 3305–3319, 2021

  2. [10]

    Tensor decompositions and applications,

    T. G. Kolda and B. W. Bader, “Tensor decompositions and applications,” SIAM Rev., vol. 51, no. 3, pp. 455–500, 2009

  3. [11]

    Tensor-train decomposition,

    I. V . Oseledets, “Tensor-train decomposition,” SIAM J. Sci. Comput. , vol. 33, no. 5, pp. 2295–2317, 2011

  4. [12]

    Tensor ring decomposition,

    Q. Zhao, G. Zhou, S. Xie, L. Zhang, and A. Cichocki, “Tensor ring decomposition,” arXiv preprint arXiv:1606.05535 , 2016

  5. [13]

    Third-order tensors as operators on matrices: A theoretical and computational framework with applications in imaging,

    M. E. Kilmer, K. Braman, N. Hao, and R. C. Hoover, “Third-order tensors as operators on matrices: A theoretical and computational framework with applications in imaging,” SIAM J. Matrix Anal. Appl. , vol. 34, no. 1, pp. 148–172, 2013

  6. [14]

    Tensor robust principal component analysis with a new tensor nuclear norm,

    C. Lu, J. Feng, Y . Chen, W. Liu, Z. Lin, and S. Yan, “Tensor robust principal component analysis with a new tensor nuclear norm,” IEEE Trans. Pattern Anal. Mach. Intell. , vol. 42, no. 4, pp. 925–938, 2019

  7. [15]

    Multi- dimensional imaging data recovery via minimizing the partial sum of tubal nuclear norm,

    T.-X. Jiang, T.-Z. Huang, X.-L. Zhao, and L.-J. Deng, “Multi- dimensional imaging data recovery via minimizing the partial sum of tubal nuclear norm,” J. Comput. Appl. Math., vol. 372, p. 112680, 2020

  8. [16]

    Weighted tensor nuclear norm minimization for tensor completion using tensor-svd,

    Y . Mu, P. Wang, L. Lu, X. Zhang, and L. Qi, “Weighted tensor nuclear norm minimization for tensor completion using tensor-svd,” Pattern Recognit. Lett., vol. 130, pp. 4–11, 2020

  9. [17]

    Laplace function based nonconvex surrogate for low-rank tensor completion,

    W.-H. Xu, X.-L. Zhao, T.-Y . Ji, J.-Q. Miao, T.-H. Ma, S. Wang, and T.- Z. Huang, “Laplace function based nonconvex surrogate for low-rank tensor completion,” Signal Process. Image Commun. , vol. 73, pp. 62– 69, 2019

  10. [18]

    Robust low-rank tensor recovery via nonconvex singular value minimization,

    L. Chen, X. Jiang, X. Liu, and Z. Zhou, “Robust low-rank tensor recovery via nonconvex singular value minimization,” IEEE Trans. Image Process., vol. 29, pp. 9044–9059, 2020

  11. [19]

    A scalable, adaptive and sound nonconvex regularizer for low-rank matrix learning,

    Y . Wang, Q. Yao, and J. Kwok, “A scalable, adaptive and sound nonconvex regularizer for low-rank matrix learning,” inProc. Web Conf., 2021, pp. 1798–1808

  12. [20]

    Computing sparse representation in a highly coherent dictionary based on difference of ℓ1 and ℓ2,

    Y . Lou, P. Yin, Q. He, and J. Xin, “Computing sparse representation in a highly coherent dictionary based on difference of ℓ1 and ℓ2,” J. Sci. Comput., vol. 64, no. 1, pp. 178–196, 2015

  13. [21]

    Minimization ofℓ1−2 for compressed sensing,

    P. Yin, Y . Lou, Q. He, and J. Xin, “Minimization ofℓ1−2 for compressed sensing,” SIAM J. Sci. Comput. , vol. 37, no. 1, pp. A536–A563, 2015

  14. [22]

    Convex- concave tensor robust principal component analysis,

    Y . Liu, B. Du, Y . Chen, L. Zhang, M. Gong, and D. Tao, “Convex- concave tensor robust principal component analysis,” Int. J. Comput. Vis., vol. 132, no. 5, pp. 1721–1747, 2024

  15. [23]

    Ratio and difference of ℓ1 and ℓ2 norms and sparse representation with coherent dictionaries,

    P. Yin, E. Esser, and J. Xin, “Ratio and difference of ℓ1 and ℓ2 norms and sparse representation with coherent dictionaries,” Commun. Inf. Syst. , vol. 14, no. 2, pp. 87–109, 2014

  16. [24]

    A scale-invariant approach for sparse signal recovery,

    Y . Rahimi, C. Wang, H. Dong, and Y . Lou, “A scale-invariant approach for sparse signal recovery,” SIAM J. Sci. Comput. , vol. 41, no. 6, pp. A3649–A3672, 2019

  17. [25]

    Accelerated schemes for the ℓ1/ℓ2 minimization,

    C. Wang, M. Yan, Y . Rahimi, and Y . Lou, “Accelerated schemes for the ℓ1/ℓ2 minimization,” IEEE Trans. Signal Process. , vol. 68, pp. 2660– 2669, 2020

  18. [26]

    Analysis and algorithms for some compressed sensing models based on ℓ1/ℓ2 minimization,

    L. Zeng, P. Yu, and T. K. Pong, “Analysis and algorithms for some compressed sensing models based on ℓ1/ℓ2 minimization,” SIAM J. Optim., vol. 31, no. 2, pp. 1576–1603, 2021

  19. [27]

    Minimization of ℓ1 over ℓ2 for sparse signal recovery with convergence guarantee,

    M. Tao, “Minimization of ℓ1 over ℓ2 for sparse signal recovery with convergence guarantee,”SIAM J. Sci. Comput., vol. 44, no. 2, pp. A770– A797, 2022

  20. [28]

    Low-rank matrix recovery problem minimizing a new ratio of two norms approximating the rank function then using an admm-type solver with applications,

    K. Gao, Z.-H. Huang, and L. Guo, “Low-rank matrix recovery problem minimizing a new ratio of two norms approximating the rank function then using an admm-type solver with applications,” J. Comput. Appl. Math., vol. 438, p. 115564, 2024

  21. [29]

    A scale-invariant relaxation in low-rank tensor recovery with an application to tensor completion,

    H. Zheng, Y . Lou, G. Tian, and C. Wang, “A scale-invariant relaxation in low-rank tensor recovery with an application to tensor completion,” SIAM Journal on Imaging Sciences , vol. 17, no. 1, pp. 756–783, 2024

  22. [30]

    Tensor robust principal component analysis via the tensor nuclear over frobenius norm,

    ——, “Tensor robust principal component analysis via the tensor nuclear over frobenius norm,” J. Sci. Comput. , vol. 104, no. 1, p. 26, 2025

  23. [31]

    Finding the largest low-rank clusters with ky fan 2- k-norm and ℓ1-norm,

    X. V . Doan and S. Vavasis, “Finding the largest low-rank clusters with ky fan 2- k-norm and ℓ1-norm,” SIAM J. Optim. , vol. 26, no. 1, pp. 274–312, 2016

  24. [32]

    Low-rank matrix recovery with ky fan 2- k-norm,

    ——, “Low-rank matrix recovery with ky fan 2- k-norm,” J. Glob. Optim., vol. 82, no. 4, pp. 727–751, 2022

  25. [33]

    Truncated ℓ1 − ℓ2 models for sparse recovery and rank minimization,

    T.-H. Ma, Y . Lou, and T.-Z. Huang, “Truncated ℓ1 − ℓ2 models for sparse recovery and rank minimization,” SIAM J. Imaging Sci. , vol. 10, no. 3, pp. 1346–1380, 2017

  26. [34]

    Multi-channel nuclear norm minus frobenius norm minimization for color image denoising,

    Y . Shan, D. Hu, Z. Wang, and T. Jia, “Multi-channel nuclear norm minus frobenius norm minimization for color image denoising,” Signal Process., vol. 207, p. 108959, 2023

  27. [35]

    Low rank matrix minimiza- tion with a truncated difference of nuclear norm and frobenius norm regularization,

    H. Guo, Q. Yu, X. Zhang, and L. Cheng, “Low rank matrix minimiza- tion with a truncated difference of nuclear norm and frobenius norm regularization,” J. Ind. Manag. Optim. , vol. 19, no. 4, 2023

  28. [36]

    Quaternion nuclear norm minus frobenius norm minimization for color image reconstruction,

    Y . Guo, G. Chen, T. Zeng, Q. Jin, and M. K.-P. Ng, “Quaternion nuclear norm minus frobenius norm minimization for color image reconstruction,” Pattern Recognit., vol. 158, p. 110986, 2025

  29. [37]

    Non-negative sparse coding,

    P. O. Hoyer, “Non-negative sparse coding,” inProc. 12th IEEE Workshop Neural Netw. Signal Process. IEEE, 2002, pp. 557–565

  30. [38]

    Comparing measures of sparsity,

    N. Hurley and S. Rickard, “Comparing measures of sparsity,” IEEE Trans. Inf. Theory, vol. 55, no. 10, pp. 4723–4741, 2009

  31. [39]

    A proximal algorithm with back- tracked extrapolation for a class of structured fractional programming,

    Q. Li, L. Shen, N. Zhang, and J. Zhou, “A proximal algorithm with back- tracked extrapolation for a class of structured fractional programming,” Appl. Comput. Harmon. Anal. , vol. 56, pp. 98–122, 2022

  32. [40]

    An equivalent reformulation and multiproximity gradient algorithms for a class of nonsmooth fractional programming,

    J. Zhou, N. Zhang, and Q. Li, “An equivalent reformulation and multiproximity gradient algorithms for a class of nonsmooth fractional programming,” Math. Oper. Res., vol. 50, no. 4, pp. 3010–3038, 2025

  33. [41]

    A min-max reformulation and proximal algorithms for a class of structured nonsmooth fractional optimization problems,

    ——, “A min-max reformulation and proximal algorithms for a class of structured nonsmooth fractional optimization problems,”J. Sci. Comput., vol. 106, no. 3, p. 72, 2026

  34. [42]

    Quaternion nuclear norms over frobenius norms minimization for robust matrix completion,

    Y . Guo, G. Chen, T. Zeng, Q. Jin, and M. K.-P. Ng, “Quaternion nuclear norms over frobenius norms minimization for robust matrix completion,” arXiv preprint arXiv:2504.21468 , 2025

  35. [43]

    Ky fan 2- k-norm model for low- rank matrix recovery with admm,

    Y . Wang, Y . Zhang, and L. Zhang, “Ky fan 2- k-norm model for low- rank matrix recovery with admm,” in Proc. 4th Int. Conf. Comput. Sci. Artif. Intell., 2020, pp. 109–114

  36. [44]

    Tensor-tensor products with invertible linear transforms,

    E. Kernfeld, M. Kilmer, and S. Aeron, “Tensor-tensor products with invertible linear transforms,” Linear Algebra Appl. , vol. 485, pp. 545– 570, 2015

  37. [45]

    Low-rank tensor completion with a new tensor nuclear norm induced by invertible linear transforms,

    C. Lu, X. Peng, and Y . Wei, “Low-rank tensor completion with a new tensor nuclear norm induced by invertible linear transforms,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. , 2019, pp. 5996–6004

  38. [46]

    R. A. Horn and C. R. Johnson, Matrix analysis. Cambridge University Press, 2012

  39. [47]

    Positive definite matrix approximation with condition number constraint,

    M. Tanaka and K. Nakata, “Positive definite matrix approximation with condition number constraint,” Optimization Letters , vol. 8, no. 3, pp. 939–947, 2014

  40. [48]

    V on neumann’s trace inequality for tensors,

    S. Chr ´etien and T. Wei, “V on neumann’s trace inequality for tensors,” Linear Algebra Appl. , vol. 482, pp. 149–157, 2015

  41. [49]

    A class of null space conditions for sparse recovery via nonconvex, non-separable minimizations,

    H. Tran and C. Webster, “A class of null space conditions for sparse recovery via nonconvex, non-separable minimizations,” Results Appl. Math., vol. 3, p. 100011, 2019

  42. [50]

    A unified alternating direction method of multipliers by majorization minimization,

    C. Lu, J. Feng, S. Yan, and Z. Lin, “A unified alternating direction method of multipliers by majorization minimization,” IEEE Trans. Pattern Anal. Mach. Intell. , vol. 40, no. 3, pp. 527–541, 2017

  43. [51]

    Proximal alternating minimization and projection methods for nonconvex problems: An approach based on the kurdyka-łojasiewicz inequality,

    H. Attouch, J. Bolte, P. Redont, and A. Soubeyran, “Proximal alternating minimization and projection methods for nonconvex problems: An approach based on the kurdyka-łojasiewicz inequality,”Math. Oper. Res., vol. 35, no. 2, pp. 438–457, 2010. 27

  44. [52]

    Exact low tubal rank tensor recovery from gaussian measurements,

    C. Lu, J. Feng, Z. Lin, and S. Yan, “Exact low tubal rank tensor recovery from gaussian measurements,” in Proc. 27th Int. Jt. Conf. Artif. Intell. (IJCAI), 2018, pp. 2504–2510

  45. [53]

    A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics,

    D. Martin, C. Fowlkes, D. Tal, and J. Malik, “A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics,” in Proc. 8th IEEE Int. Conf. Comput. Vis. (ICCV) , vol. 2. IEEE, 2001, pp. 416–423

  46. [54]

    Limited-angle ct reconstruc- tion via the ℓ1/ℓ2 minimization,

    C. Wang, M. Tao, J. G. Nagy, and Y . Lou, “Limited-angle ct reconstruc- tion via the ℓ1/ℓ2 minimization,” SIAM J. Imaging Sci. , vol. 14, no. 2, pp. 749–777, 2021

  47. [55]

    The case of equality in the von neumann trace inequality,

    D. Rhea, “The case of equality in the von neumann trace inequality,” Preprint, 2011

  48. [56]

    R. T. Rockafellar, Convex analysis. Princeton University Press, 1997, vol. 28

  49. [57]

    Perturbation bounds in connection with singular value decomposition,

    P.- ˚A. Wedin, “Perturbation bounds in connection with singular value decomposition,” BIT Numer. Math., vol. 12, no. 1, pp. 99–111, 1972

Pith tools

Reviewed June 26, 2026 · model on record in the stance chip above.