Pith. sign in

REVIEW 1 major objections 6 minor 34 references

Quadratic Surface Support Vector Machine with L1 Norm Regularization

T0 review · 1 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read Adding an $\ell_1$ penalty to quadratic surface SVMs is claimed to yield linear behavior at large penalties and sparse recovery on quadratic data.

desk verdict A useful L1-regularized QSSVM formulation with a correct reduction-to-SVM theorem, but the advertised exact sparsity-recovery result rests on a KKT verification that fails on the support; still worth a serious referee. read the letter →

arxiv 1908.08616 v2 pith:5HRMFDVW submitted 2019-08-22 cs.LG math.OCstat.ML

classification cs.LGmath.OCstat.ML MSC 68T0590C2090C2562H30
keywords binaryclassificationkernel-freequadraticsurfaceSVML1regularizationsparsitypatternrecoveryconvexprogrammingsoftmargincurvaturecontrolsupportvectormachine
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 two kernel-free classifiers, L1-QSSVM and its soft-margin version L1-SQSSVM, that add an $\ell_1$ penalty on the Hessian coefficients of the quadratic separating surface. The central claim is that this single term turns the quadratic-surface SVM into a spectrum of models: small penalties reproduce the known quadratic surface behavior, while large penalties force the separating surface to collapse onto the standard linear SVM whenever the data are (almost) linearly separable. For quadratically separable data whose generating quadratic form is sparse, the paper further claims that sufficiently large penalties recover the true zero pattern of that form and drive the slack variables to zero. These properties matter because the unregularized QSSVM family has no built-in reason to return a hyperplane on linear data, and because an explicit curvature penalty gives practitioners a principled way to encode prior beliefs about the shape of the decision boundary.

What carries the argument

The central object is the regularized quadratic program that minimizes, over $z=(\mathrm{hvec}(W),b)$, $c$, and $\xi$, the objective $\frac{1}{2}z^\top G z + \lambda\|Vz\|_1 + \mu\sum_i \xi_i$ subject to $y_i(z^\top r_i+c)\ge 1-\xi_i$, where $G=2\sum_i (H^{(i)})^\top H^{(i)}$ encodes the summed squared norm of $Wx_i+b$ and $V$ picks out the Hessian coefficients. The $\ell_1$ term is the mechanism: its subdifferential enters the KKT stationarity condition, and at any coordinate with $w_j\neq 0$ the condition requires $\lambda\,\mathrm{sign}(w_j)$ to be balanced by the other gradient terms. A large $\lambda$ therefore cannot be absorbed by a nonzero coefficient unless the remaining terms oppose it, which is how the separating surface is flattened to a hyperplane and how the paper claims zero patterns of a sparse generating $W$ are selected. Positive definiteness of $G$, proven under full-column-rank of $X$ and the all-ones vector not lying in its column space, is what makes $z$ unique.

What would settle it

On a dataset generated by a known sparse $W$, compute a solution of the restricted QSSVM that fixes the true zero coordinates to zero, read off its multipliers $\beta^*$, and solve L1-QSSVM with $\lambda>\|\beta^*\|_\infty$; if the returned $w$ has any nonzero entry outside the true support, the KKT-transfer step in the sparsity-recovery argument fails. The paper's own 10-feature synthetic example with an eight-entry banded $W$ provides a ready-made instance for this check.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes the following. The matrix $G=2\sum_i (H^{(i)})^\top H^{(i)}$ is positive definite for almost every data matrix $X\in\mathbb{R}^{m\times n}$ with $m\ge n+1$ (provided the all-ones vector is not in the column space of $X$), hence the decision-variable vector $z=(\mathrm{hvec}(W),b)$ is unique for almost all data sets. If the data are linearly separable, the pair $W=0$, $b=u^*$, $c=d^*$ obtained from the hard-margin SVM becomes a solution of L1-QSSVM once $\lambda$ exceeds an explicit bound; the same holds for the soft-margin pair from SSVM relative to L1-SQSSVM. If the data are quadratically separable by a sparse $W$, the paper argues that solving the $\ell_1$-regularized problem with $\lambda$ above the infinity norm of the restricted-problem multipliers recovers the support of $W$ exactly, with $\xi=0$ once $\mu$ is also large.

Load-bearing premise

The sparsity-recovery claim rests on assuming a solution of the problem with the true zero coordinates forced to zero automatically satisfies the optimality conditions of the full $\ell_1$ problem once $\lambda$ is large; that transfer fails on every coordinate where the recovered coefficient is nonzero, because the $\ell_1$ condition would demand $\lambda$ times the sign of that coefficient be zero.

Editorial extensions

If this is right

  • For linearly separable data, L1-QSSVM and L1-SQSSVM reduce to the standard hard- and soft-margin SVM solutions at finite, computable $\lambda$ thresholds.
  • The $\lambda$ parameter acts as a curvature dial, covering the parental QSSVM surface at $\lambda=0$ and flat hyperplanes at large $\lambda$, so a single model family spans both regimes without switching kernels.
  • For quadratically separable data with a sparse generator, the model selects feature interactions automatically, provided the sparsity-recovery argument holds as stated.
  • The soft-margin version is feasible for any data set, so practical classification does not require an a priori separability check; uniqueness of $z$ fails only on a measure-zero set of data matrices.
  • With large enough $\mu$ on separable data, the slack variables vanish, so the soft-margin model returns a hard separating surface.

Reading between the lines

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

  • If the KKT gap in the sparsity theorem is repaired, L1-SQSSVM would amount to a kernel-free analogue of sparse feature-interaction learning with no kernel selection step; the numerical support plots suggest a weaker near-support containment may already hold empirically.
  • A natural extension is to weight the $\ell_1$ penalty separately for diagonal versus off-diagonal Hessian entries, since diagonal terms represent single-feature curvature and off-diagonal terms represent pairwise interactions, perhaps using group sparsity on the interaction block.
  • The uniqueness of $z$ for almost every data matrix suggests that support recovery could be analysed under a restricted-isometry-type condition on the lifted vectors $r^{(i)}$, connecting to compressed-sensing-style guarantees.
  • Robustness to label noise is not addressed; a testable extension is to perturb a small fraction of labels in the benchmark sets and measure whether the $\lambda$ threshold for flatness shifts substantially.
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

1 major / 6 minor

Summary. The paper introduces l1-norm regularized versions of the kernel-free quadratic surface SVM, called L1-QSSVM and its soft-margin variant L1-SQSSVM. The main theoretical claims are: (i) an optimal solution always exists and the z-component is unique for almost all data matrices; (ii) on linearly separable data, for sufficiently large lambda, the l1-regularized models reduce to the standard hard- or soft-margin SVM; and (iii) on quadratically separable data generated by a sparse matrix W, for large enough lambda and mu, the model recovers the true sparsity pattern of W with vanishing slack. The paper also presents numerical experiments on synthetic and UCI benchmark data sets. The existence, z-uniqueness, generic positive definiteness of G, and finite-lambda SVM-reduction results are argued correctly. However, the sparsity-pattern recovery claim rests on a KKT verification that is algebraically incorrect on the support of the restricted solution, so the flagship theoretical result is not established.

Significance. If the sparsity-recovery theorem were valid, the paper would provide a useful contribution: a kernel-free quadratic SVM whose l1 term both controls curvature and performs feature-interaction selection, with a finite-lambda bridge to linear SVM. The paper's strongest verified results are the solution existence (Theorem 4.1), generic uniqueness of z via positive definiteness of G (Theorems 4.2-4.4), vanishing margin for large mu (Theorem 4.5 in its z and xi aspects), and finite-lambda reduction to SVM (Theorem 4.6 and Corollary 4.6.1); these are supported by direct convex-analysis arguments rather than by fitting parameters to data. The numerical study is extensive and shows accuracy gains on several benchmark data sets. However, because the sparsity-recovery proof fails exactly at the point that distinguishes the paper from prior QSSVM work, the paper's central theoretical promise is currently unsupported, and the experiments do not quantify exact support recovery. If repaired, the result would be a meaningful advance; in its current form the paper cannot be recommended.

major comments (1)
  1. [Section 4.3 (Corollary 4.6.2)] The KKT verification for sparsity-pattern recovery fails on the support coordinates. Let (w*, b*, c*, alpha*, beta*) be a KKT point of the restricted problem (R-QSSVM''), with beta* extended by zeros outside the zero set Z. For any j not in Z with w*_j nonzero, the first stationarity equation of L1-QSSVM in (16) requires 2(sum_i M^(i)T M^(i) w* + sum_i M^(i)T b*)_j - sum_i alpha*_i y^(i) s^(i)_j + lambda sign(w*_j) = 0, because the l1 subgradient is the singleton {sign(w*_j)}. The restricted KKT equation gives the same expression without the lambda term equal to zero, since beta*_j = 0 off Z. Substituting yields lambda sign(w*_j) = 0, which is impossible for lambda > 0. The condition lambda > ||beta*||_infinity only controls coordinates in Z, where w*_j = 0 and the subgradient interval [-1,1] is available; it says nothing about nonzero coordinates outside Z. Thus the constructed point is not a KKT point of L1-QSSVM, and the sparsity-recovery theorem and Corollary 4.6.2 are unproven. The argument is not actually 'similar to Theorem 4.6', because in Theorem 4.6 the candidate has w* = 0, making the l1 subdifferential an interval.
minor comments (6)
  1. [Theorem 4.1 proof] The sentence 'Lemma 2.2 implies that (L1-SQSSVM') has an optimal solution with a finite objective value and so does (L1-SQSSVM')' repeats the same model name; the second occurrence should refer to (L1-QSSVM').
  2. [Theorems 4.2 and 4.3] Theorem 4.2 refers to 'G defined in (19)', but G is first defined in equation (7); equation (19) is a later restatement and should not be the primary reference.
  3. [KKT systems (10), (12), (14), (16)] The multiplier conditions are written with strict positivity, e.g., alpha*_i > 0 and beta*_i > 0. The correct condition is nonnegativity; strict positivity can fail for non-support vectors, and the proofs only need nonnegativity.
  4. [Theorem 4.5] The uniqueness of c* is asserted by reference to Theorem 5 of [21], but c does not appear in the objective of (L1-SQSSVM'), and after fixing z*, the pair (c*, xi*) solves the linear program (9), which need not have a unique c when margin constraints are satisfied with slack. The statement should either prove uniqueness of c* or be weakened to uniqueness of (z*, xi*).
  5. [Figure 5 and sparsity experiments] The experiment on sparsity-pattern detection would be much more informative with a quantitative support-recovery measure, such as false positives and false negatives relative to the support of W in (18); the displayed heatmaps alone do not verify exact recovery.
  6. [Section 5, parameter grid] The notation 'log2 mu in {-3.-2,...,20}' should read '{-3,-2,...,20}'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the theoretical claims are derived from the model definitions through standard KKT and convex-analysis arguments, and no load-bearing step reduces to its own inputs.

full rationale

The paper derives its theoretical claims (solution existence, z-uniqueness, equivalence with SVM/SSVM, vanishing margin, and sparsity recovery) directly from the proposed models via convex analysis and KKT conditions. The candidate solutions in Theorems 4.5 and 4.6 are constructed from the corresponding hard-margin problems and then verified against the KKT systems (12) and (16); the verification is algebraic and does not presuppose the theorem's conclusion. The sparsity-recovery argument compares the restricted problem R-QSSVM'' with L1-QSSVM through stationarity, requiring λ to exceed ‖β*‖∞; this is an existential condition on the penalty parameter, not a fitted value or a quantity derived from the claimed sparsity pattern. Citations to prior work are mostly external (e.g., [21], [24], [3], [17]), and the few self-citations (e.g., [12], [26], [27], [32], [33]) are background references on l1 regularization or applications and are not load-bearing for the main theorems. The apparent flaw in the sparsity-recovery KKT verification on support coordinates is a mathematical correctness concern, not circularity: the claimed conclusion is not equivalent to its input by construction. No fitted parameters are renamed as predictions, and no load-bearing uniqueness theorem is imported from the authors' own prior work. Therefore no circular step is identified.

Assumptions & free parameters 2 free parameters · 6 assumptions · 0 invented entities

The central claims rest on generic full-rank conditions on the data matrix, separability assumptions for the reduction and sparsity theorems, and standard convex-analysis results (KKT sufficiency, bounded Lagrange multipliers under Slater's condition). No new entities are postulated. The hyperparameters lambda and mu are user-chosen penalties, tuned on data in the experiments, and are not fitted to derive any theorem.

free parameters (2)
  • lambda (L1 penalty) = tuned by grid search in experiments; values up to 2^25 in the sparsity experiment
    Controls the L1 penalty on W and is central to the reduction and sparsity theorems; in experiments it is tuned on validation accuracy, so it is a data-fitted hyperparameter.
  • mu (misclassification penalty) = tuned via SQSSVM grid search with log2 mu in the range -3 to 20
    Penalty for slack variables in L1-SQSSVM; tuned on data before running the model in experiments.
assumptions (6)
  • domain assumption X has full column rank (A1)
    Invoked in Theorem 4.3 to prove G is positive definite; the paper argues it is mild since m is typically much larger than n.
  • domain assumption 1_m is not in the column space of X (A2)
    Needed together with (A1) for G to be positive definite; Theorem 4.4 shows the complement has measure zero for m at least n+1.
  • domain assumption Quadratic separability (5) for Theorems 4.5 and the sparsity results
    Guarantees feasibility of the hard-margin problem and Slater's condition so Lagrange multipliers are bounded.
  • domain assumption Linear separability (6) for Theorem 4.6
    Ensures the standard SVM has a strictly feasible solution and bounded multipliers, used to bound the required lambda.
  • standard math KKT conditions are necessary and sufficient for the convex programs
    Used throughout Theorems 4.5, 4.6, and the sparsity argument; standard for convex QPs under Slater's condition.
  • standard math Boundedness of Lagrange multipliers under Slater's condition (Exercise 5.3.1 in [3])
    Used to bound the norm of alpha* in Theorem 4.5 and the norm of beta* in Theorem 4.6.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quadratic Surface Support Vector Machine with L1 Norm Regularization." pith.science (2026). https://pith.science/paper/5HRMFDVW

@misc{pith2026190808616,
  author       = {Pith},
  title        = {Pith review of: Quadratic Surface Support Vector Machine with L1 Norm Regularization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5HRMFDVW}},
  note         = {Machine review of arXiv:1908.08616}
}
abstract

We propose $\ell_1$ norm regularized quadratic surface support vector machine models for binary classification in supervised learning. We establish their desired theoretical properties, including the existence and uniqueness of the optimal solution, reduction to the standard SVMs over (almost) linearly separable data sets, and detection of true sparsity pattern over (almost) quadratically separable data sets if the penalty parameter of $\ell_1$ norm is large enough. We also demonstrate their promising practical efficiency by conducting various numerical experiments on both synthetic and publicly available benchmark data sets.

Figures

Figures reproduced from arXiv: 1908.08616 by the authors.

Figure 1
Figure 1. L1-QSSVM performance on linearly and quadratically separable data sets. [PITH_FULL_IMAGE:figures/full_fig_p016_1.png] view at source ↗
Figure 2
Figure 2. Influence of the parameter λ on the curvature of the optimal solution of L1-SQSSVM. (a) Pm i=1ξi against parameter µ. (Artificial 3-D data) (b) Relative error against parameter µ. (Artificial 3-D data) [PITH_FULL_IMAGE:figures/full_fig_p017_2.png] view at source ↗
Figure 3
Figure 3. Influence of the parameter µ on the behavior of the optimal solution of L1-SQSSVM. Next, we numerically demonstrate that the `1 norm term in our proposed models is significant in classification, namely, a suitable parameter λ > 0 exists that leads to a better performance than when λ = 0. We only focus on the soft margin model because both models resemble similar behavior in this sense. To show that the existence of … view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Accuracy score against the parameter λ Consider the case where a given data set is quadratically separable with a sparse W matrix, i.e., when the separation surface f(x) = 1 2 x TWx+b T x+c = 0 has a sparse W. By applying L1-SQSSVM one can see that the `1 norm regulari…
Figure 5
Figure 5. Figure 5: Sparsity pattern detection using L1-SQSSVM as parameter [PITH_FULL_IMAGE:figures/full_fig_p019_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 30 canonical work pages

  1. [21]

    Softquadraticsurfacesupportvector machine for binary classification.Asia-Pacific Journal of Operational Research, 33(06):1650046, 2016

    JianLuo, Shu-CherngFang, ZhibinDeng, andXiaolingGuo. Softquadraticsurfacesupportvector machine for binary classification.Asia-Pacific Journal of Operational Research, 33(06):1650046, 2016

  2. [1]

    UCI machine learning repository, 2007

    Arthur Asuncion and David Newman. UCI machine learning repository, 2007

  3. [2]

    Quadratickernel-freeleastsquaressupportvector machine for target diseases classification.Journal of Combinatorial Optimization, 30(4):850–870, 2015

    YanqinBai, XiaoHan, TongChen, andHuaYu. Quadratickernel-freeleastsquaressupportvector machine for target diseases classification.Journal of Combinatorial Optimization, 30(4):850–870, 2015

  4. [3]

    Nonlinear programming

    Dimitri P Bertsekas. Nonlinear programming. Journal of the Operational Research Society, 48(3):334–334, 1997

  5. [4]

    Springer Science & Business Media, 2010

    Jonathan Borwein and Adrian S Lewis.Convex analysis and nonlinear optimization: theory and examples. Springer Science & Business Media, 2010

  6. [5]

    Support-vector networks.Machine learning, 20(3):273–297, 1995

    Corinna Cortes and Vladimir Vapnik. Support-vector networks.Machine learning, 20(3):273–297, 1995

  7. [6]

    Cambridge university press, 2000

    Nello Cristianini and John Shawe-Taylor.An introduction to support vector machines and other kernel-based learning methods. Cambridge university press, 2000

  8. [7]

    Quadratic kernel-free non-linear support vector machine.Journal of Global Opti- mization, 41(1):15–30, 2008

    Issam Dagher. Quadratic kernel-free non-linear support vector machine.Journal of Global Opti- mization, 41(1):15–30, 2008

Show all 34 references
  1. [8]

    A generalized approach to sparse and stable portfolio optimization problem

    Zhifeng Dai and Fenghua Wen. A generalized approach to sparse and stable portfolio optimization problem. Journal of Industrial and Management Optimization, 14(4):1651–1666, 2018

  2. [9]

    Chapman and Hall/CRC, 2012

    Naiyang Deng, Yingjie Tian, and Chunhua Zhang.Support vector machines: optimization based theory, algorithms, and extensions. Chapman and Hall/CRC, 2012

  3. [10]

    Asurveyofmachinelearninginwirelesssensornetworksfromnetworking and application perspectives

    MaDiandErMengJoo. Asurveyofmachinelearninginwirelesssensornetworksfromnetworking and application perspectives. In2007 6th international conference on information, communica- tions & signal processing, pages 1–5. IEEE, 2007

  4. [11]

    Schur complements and applications

    Jean Gallier. Schur complements and applications. InGeometric Methods and Applications, pages 431–437. Springer, 2011

  5. [12]

    A kernel-free double well po- tential support vector machine with applications

    Zheming Gao, Shu-Cherng Fang, Jian Luo, and Negash Medhin. A kernel-free double well po- tential support vector machine with applications. European Journal of Operational Research, 2020

  6. [13]

    Rescaled pure greedy algorithm for convex optimization

    Zheming Gao and Guergana Petrova. Rescaled pure greedy algorithm for convex optimization. Calcolo, 56(2):15, 2019

  7. [14]

    High dimensional data classification and feature selec- tion using support vector machines.European Journal of Operational Research, 265(3):993–1004, 2018

    Bissan Ghaddar and Joe Naoum-Sawaya. High dimensional data classification and feature selec- tion using support vector machines.European Journal of Operational Research, 265(3):993–1004, 2018

  8. [15]

    A new method on gene selection for tissue classification.Journal of Industrial and Management Optimization, 3(4):739, 2007

    Ying Hao and Fanwen Meng. A new method on gene selection for tissue classification.Journal of Industrial and Management Optimization, 3(4):739, 2007

  9. [16]

    Complexity measures of supervised classification problems.IEEE Transactions on Pattern Analysis & Machine Intelligence, (3):289–300, 2002

    Tin Kam Ho and Mitra Basu. Complexity measures of supervised classification problems.IEEE Transactions on Pattern Analysis & Machine Intelligence, (3):289–300, 2002. 26

  10. [17]

    Solution existence and stability of quadrat- ically constrained convex quadratic programs.Optimization Letters, 6(2):363–373, 2012

    DS Kim, Nguyen Nang Tam, and Nguyen Dong Yen. Solution existence and stability of quadrat- ically constrained convex quadratic programs.Optimization Letters, 6(2):363–373, 2012

  11. [18]

    Applications of machine learning and rule induction.Com- munications of the ACM, 38(11):54–64, 1995

    Pat Langley and Herbert A Simon. Applications of machine learning and rule induction.Com- munications of the ACM, 38(11):54–64, 1995

  12. [19]

    Taking advantage of sparsity in multi-task learning.arXiv preprint arXiv:0903.1468, 2009

    Karim Lounici, Massimiliano Pontil, Alexandre B Tsybakov, and Sara Van De Geer. Taking advantage of sparsity in multi-task learning.arXiv preprint arXiv:0903.1468, 2009

  13. [20]

    Fuzzy quadratic surface support vector machine based on Fisher discriminant analysis.Journal of Industrial and Management Optimization, 12(1):357–373, 2016

    Jian Luo, Shu-Cherng Fang, Yanqin Bai, and Zhibin Deng. Fuzzy quadratic surface support vector machine based on Fisher discriminant analysis.Journal of Industrial and Management Optimization, 12(1):357–373, 2016

  14. [22]

    Benchmarking robustness of load forecasting models under data integrity attacks.International Journal of Forecasting, 34(1):89–104, 2018

    Jian Luo, Tao Hong, and Shu-Cherng Fang. Benchmarking robustness of load forecasting models under data integrity attacks.International Journal of Forecasting, 34(1):89–104, 2018

  15. [23]

    The elimination matrix: some lemmas and applications.SIAM Journal on Algebraic Discrete Methods, 1(4):422–449, 1980

    Jan R Magnus and H Neudecker. The elimination matrix: some lemmas and applications.SIAM Journal on Algebraic Discrete Methods, 1(4):422–449, 1980

  16. [24]

    Uniqueness of solution in linear programming.Linear Algebra and its Applica- tions, 25:151–162, 1979

    OL Mangasarian. Uniqueness of solution in linear programming.Linear Algebra and its Applica- tions, 25:151–162, 1979

  17. [25]

    Machine learning approaches to manufacturing.CIRP annals, 45(2):675–712, 1996

    László Monostori, András Márkus, Hendrik Van Brussel, and E Westkämpfer. Machine learning approaches to manufacturing.CIRP annals, 45(2):675–712, 1996

  18. [26]

    A survey on compressive sensing: clas- sical results and recent advancements.Journal of Mathematical Modeling, 8(3):309–344, 2020

    Ahmad Mousavi, Mehdi Rezaee, and Ramin Ayanzadeh. A survey on compressive sensing: clas- sical results and recent advancements.Journal of Mathematical Modeling, 8(3):309–344, 2020

  19. [27]

    Solution uniqueness of convex piecewise affine functions based optimization with applications to constrained𝓁1 minimization

    Seyedahmad Mousavi and Jinglai Shen. Solution uniqueness of convex piecewise affine functions based optimization with applications to constrained𝓁1 minimization. ESAIM: Control, Optimi- sation and Calculus of Variations, 25:56, 2019

  20. [28]

    Pedregosa, G

    F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Pret- tenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay. Scikit-learn: Machine learning in Python.Journal of Machine Learnin...

  21. [29]

    A fast 𝓁1-solver and its applications to robust face recognition.Journal of Industrial and Manage- ment Optimization, 8:163–178, 2012

    Huining Qiu, Xiaoming Chen, Wanquan Liu, Guanglu Zhou, Yiju Wang, and Jianhuang Lai. A fast 𝓁1-solver and its applications to robust face recognition.Journal of Industrial and Manage- ment Optimization, 8:163–178, 2012

  22. [30]

    Stable sparse approximations via nonconvex optimization

    Rayan Saab, Rick Chartrand, and Ozgur Yilmaz. Stable sparse approximations via nonconvex optimization. In2008 IEEE International Conference on Acoustics, Speech and Signal Processing, pages 3885–3888. IEEE, 2008

  23. [31]

    Learning with kernels: support vector machines, regularization, optimization, and beyond

    Bernhard Scholkopf and Alexander J Smola. Learning with kernels: support vector machines, regularization, optimization, and beyond. MIT press, 2001

  24. [32]

    Least sparsity ofp-norm based optimization problems with p> 1

    Jinglai Shen and Seyedahmad Mousavi. Least sparsity ofp-norm based optimization problems with p> 1. SIAM Journal on Optimization, 28(3):2721–2751, 2018

  25. [33]

    Exact support and vector recovery of constrained sparse vectors via constrained matching pursuit.arXiv preprint arXiv:1903.07236, 2019

    Jinglai Shen and Seyedahmad Mousavi. Exact support and vector recovery of constrained sparse vectors via constrained matching pursuit.arXiv preprint arXiv:1903.07236, 2019

  26. [34]

    Robust and sparse portfolio model for index tracking

    Chao Zhang, Jingjing Wang, and Naihua Xiu. Robust and sparse portfolio model for index tracking. Journal of Industrial and Management Optimization, 15(3):1001–1015, 2019. Received xxxx 20xx; revised xxxx 20xx. 27

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.