Pith. sign in

REVIEW 4 major objections 4 minor 62 references

Testing for large-dimensional covariance matrix under differential privacy

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper claims that testing whether a large covariance matrix equals the identity can be made differentially private by adding Laplace noise to sample eigenvalues, with no bounded-data assumption, while staying asymptotically…

desk verdict First DP test for high-dimensional covariance structure with a real sensitivity bound, but the two-stage noise calibration in Algorithm 1 breaks the advertised privacy guarantee. read the letter →

arxiv 2506.02410 v1 pith:OKDOIER5 submitted 2025-06-03 stat.ME

Shiwei Sang , Yicheng Zeng , Xuehu Zhu , Shurong Zheng This is my paper
classification stat.ME MSC 62H1562H1062F0360B2068P27
keywords differentialprivacyrandommatrixtheorylikelihoodratiostatisticscovariancetestinghigh-dimensionaldatasub-Gaussiandistributioneigenvaluesensitivitylocalalternatives
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 proposes the first differentially private test for the high-dimensional null hypothesis $H_0: \Sigma = I_d$, where the dimension $d$ and the sample size $n$ grow together with $d/n \to y \in (0,\infty)$. The central move is to query only the sample eigenvalues, perturb each one with Laplace noise calibrated by a new high-probability bound on the $\ell_1$ global sensitivity of the eigenvalue map, valid for unbounded sub-Gaussian data. The privatized statistic is the maximum of three eigenvalue-based loss statistics — an entropy-like loss, a quadratic loss, and an absolute-deviation loss — and it is shown to be asymptotically distribution-free under the null, with critical values from a known Gaussian maximum distribution, and to detect local alternatives at the rate $1/\sqrt{n}$. If correct, a data curator can test $\Sigma = I_d$ without releasing eigenvectors or raw eigenvalues and, with high probability, without any bounded-data assumption.

What carries the argument

The proof rests on two pillars. The first is the sensitivity analysis: using the tail bound for quadratic forms of sub-Gaussian vectors of Hsu et al. (2012), the paper shows that replacing one row of the design matrix changes the $\ell_1$ distance between the eigenvalue vectors by at most $(x_1^\top x_1 + \tilde{x}_1^\top \tilde{x}_1)/n$, and bounds this quantity in probability. The second is a DP version of the classical CLT for linear spectral statistics: eigenvalue rigidity lets the authors replace each $\lambda_i$ by its classical location $\alpha_i$ at a cost of $o_p(K^{-1/2})$, after which a Lyapunov CLT applies to $g_m(\alpha_i + \ell_i)$ because the independent Laplace noise $\ell_i$ provides a non-degenerate variance. The mean and covariance terms are then evaluated as Riemann sums against the generalized Marcenko-Pastur density. The Laplace noise, of order $d/n$, dominates the $O(K^{-1})$ fluctuations of the sample eigenvalues, so the asymptotic variance is non-zero, is governed by the noise, and does not depend on the unknown covariance under the null, which is what makes the test distribution-free.

What would settle it

Run Algorithm 1 on data whose sub-Gaussian parameter is $\sigma = 2$ (for example, standardized two-point or narrowly supported sub-Gaussian vectors) while the algorithm assumes $\sigma = 1$ and takes $\tilde{\gamma}$ equal to the true $\gamma$. If the empirical Type I error at $\varepsilon = 8$ diverges from the nominal level as $n$ grows, or if a finite-sample attack on the released eigenvalues can detect a single changed row with probability exceeding the $\varepsilon$-DP bound, the sensitivity calibration premise is false.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that spectral tests for covariance structure can retain their asymptotic properties after differential privacy in the proportional-growth regime $d/n \to y \in (0,\infty)$. The central discovery is that the $\ell_1$ sensitivity of the sample eigenvalue query is bounded, with probability at least $1 - 2e^{-nt}$, by $\sigma^2\{2\mathrm{Tr}(\Sigma)/n + 4\sqrt{\mathrm{Tr}(\Sigma^2)}\sqrt{t/n} + 4\|\Sigma\| t\}$, where $\sigma$ is the sub-Gaussian parameter of the standardized population. Taking $t = n^{-2r}$ yields the simplified scale $2.01\sigma^2\gamma d/n$, so Laplace noise at this rate makes the released eigenvalues $\varepsilon$-DP with probability at least $1 - 2\exp\{-n^{1-2r}\}$. For the three loss functions $g_1(x)=|x|-\log|x|-1$, $g_2(x)=|x-1|^2$, and $g_3(x)=|x-1|$, the perturbed linear spectral statistics $\frac{1}{K}\sum_{i=1}^K g_m(\lambda_i + \ell_i)$ are jointly Gaussian under $H_0$, with mean and covariance given by convolving the limiting spectral measure with the Laplace density, so $T^{\mathrm{dp}}_{\max}$ is asymptotically distribution-free; the same CLT implies power tends to one whenever any coordinate's mean shift exceeds $\omega(1/\sqrt{K})$, giving detection at the $1/\sqrt{n}$ rate.

Load-bearing premise

The load-bearing premise is that the sub-Gaussian parameter of the standardized data equals exactly $1$ and that the user's preset bound $\tilde{\gamma}$ truly satisfies $\tilde{\gamma} \ge \mathrm{Tr}(\Sigma)/d$; if either fails, the Laplace scale is too small, the high-probability $\varepsilon$-DP guarantee collapses, and the null variance used for critical values is misspecified.

Editorial extensions

If this is right

  • A data curator can release privatized eigenvalues, and hence a test decision, for $\Sigma = I_d$ while keeping the data $\varepsilon$-DP with probability at least $1 - 2\exp\{-n^{1-2r}\}$.
  • The test is asymptotically distribution-free: critical values come from the distribution of the maximum of three correlated Gaussians, so no Monte Carlo or resampling is needed to set the threshold.
  • Power tends to one for any alternative whose mean shift in at least one of the three losses is of order $\omega(1/\sqrt{n})$, and the test retains non-trivial power at the $1/\sqrt{n}$ boundary.
  • Because the released objects are only the noisy eigenvalues, the eigenvectors of the sample covariance matrix are never exposed, so the covariance matrix itself cannot be reconstructed.
  • The noise-inflated variance is the leading term in the fluctuation of the privatized statistic, so the null distribution does not depend on the unknown $\Sigma$ under $H_0$ beyond the limit ratio $y$.

Reading between the lines

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

  • The privacy guarantee is conditional on premises the user cannot verify privately: if the true sub-Gaussian parameter $\sigma$ exceeds $1$ or the true trace ratio $\gamma$ exceeds the preset $\tilde{\gamma}$, the injected noise is silently too small and the claimed $\varepsilon$-DP breaks; a two-stage procedure that spends part of the privacy budget to estimate the scale would be a natural extensi
  • Because the required noise scale is of order $d/n$, the method is confined to the proportional-growth regime; the authors' own concluding remark suggests that as $d/n \to \infty$ the same statistics cannot be privatized at this rate, which points toward an information-theoretic price of privacy in ultra-high-dimensional spectral testing worth investigating.
  • The maximum-of-three integrated statistic offers a template for private high-dimensional testing in general — combining losses with different sensitivity and power profiles can rescue detection when the signal shape is unknown — and the same integration could be ported to sphericity, independence, or two-sample covariance tests.
  • Releasing privatized eigenvalues rather than a single test decision enables multiple downstream analyses, but the paper does not analyze composition of many queries; repeated releases would require advanced composition or an explicit privacy-loss budget.
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

4 major / 4 minor

Summary. The paper proposes differentially private tests for the high-dimensional hypothesis H0: Sigma = I_d when d/n -> y in (0, inf). The test statistics (1.4)-(1.6) are built from Laplace-perturbed sample eigenvalues, with the noise scale derived from a high-probability bound on the l1 sensitivity of the eigenvalue map for sub-Gaussian data (Theorem 1 and Theorem 2). The authors claim an asymptotically distribution-free null limit (Theorem 3), consistency against local alternatives at rate 1/sqrt(n) (Section 3.3), and support this with simulations and a real-data example. The statistical core is an original combination of random matrix theory and DP, but the privacy guarantee as implemented in Algorithm 1 is not established, and the null CLT is proved only under a support-separation condition that fails at y=1 and y>1.

Significance. If the results were correct as stated, this would be a valuable first contribution to DP-based testing of large-dimensional covariance structures, and the sensitivity analysis for unbounded sub-Gaussian data is a useful technical step. The paper is also commendable for stating detailed proofs of the sensitivity bound and of the CLT, and for providing extensive simulations. However, the central privacy claim is not supported by the algorithm actually implemented: the noise scale is data-dependent in a way that is not analyzed, and the simulations include a model whose sub-Gaussian parameter is not the assumed sigma=1. The asymptotic null distribution is also unproved for the important cases y=1 and y>1. These are load-bearing issues for the paper's main advertised contributions.

major comments (4)
  1. [Algorithm 1, Steps 3-4; Theorem 2] Theorem 2 proves a privacy guarantee for an oracle mechanism that adds Laplace noise with scale 2.01*sigma^2*gamma*d/(n*epsilon), where gamma=Tr(Sigma)/d. Algorithm 1 instead uses the data-dependent estimate hat(gamma) = d^{-1}|sum_i (lambda_i + ell*_i)| from Step 3 to set the scale in Step 4. Under H0, gamma=1 and hat(gamma)=1+O_p(K^{-1/2}) with a symmetric centered distribution, so P(hat(gamma)<1) tends to 1/2. Whenever hat(gamma)<gamma, the Laplace scale is strictly smaller than the sensitivity bound in Theorem 2, and for neighboring datasets whose sensitivity attains the bound, the output density ratio can exceed e^epsilon. Thus the mechanism in Algorithm 1 is not epsilon-DP with the probability claimed in Theorem 2. No composition argument or adaptive-sensitivity analysis is provided for the two-stage release of hat(gamma) and the subsequent eigenvalues.
  2. [Remark 3 and Model II in Section 4.1] The algorithm's privacy guarantee relies on the unverified assumptions sigma=1 and gamma <= gamma_tilde. In the simulations, Model II generates z from Unif([-sqrt(3), sqrt(3)]^d), which is sub-Gaussian with sigma > 1 (Hoeffding's lemma gives sigma=sqrt(3) up to constants). With gamma_tilde=2, the scale used in Algorithm 1 is smaller than 2.01*sigma^2*gamma*d/(n*epsilon) under H0, so the privacy claim in Theorem 2 does not cover the reported Model II results. The paper should either treat sigma as a known upper bound and use it in the scale, or provide a privacy-preserving procedure to estimate or conservatively bound sigma; the current Remark 3 simply assumes the problem away.
  3. [Theorem 3 and Lemma 5/Assumption 4] Theorem 3 is stated under only Assumptions 1-2 for all y in (0,inf), but its proof invokes Lemma 5, whose proof requires Assumption 4, in particular that the support of the limiting spectral distribution is bounded away from zero. Under H0 with y=1, the Marcenko-Pastur left edge is 0, and for y>1 the support includes 0. Hence Assumption 4 fails exactly in the regimes y=1 and y>1 that are included in the theorem's statement and in the simulations (y=1 and y=5). The asymptotic distribution-free null limit is therefore not proved for these cases; a separate treatment of the zero edge, or a restricted statement of Theorem 3, is needed.
  4. [Theorem 2 and Definition 1] The statement that the privatized statistics are 'epsilon-DP with probability at least 1 - 2exp(-n^{1-2r})' is not the standard worst-case guarantee of Definition 1. The probability is over the random data, whereas DP requires the inequality to hold for every fixed neighboring pair. If the intended guarantee is (epsilon,delta)-DP with delta=2exp(-n^{1-2r}), the proof must spell out the conditioning on the sensitivity event and how the failure event is handled uniformly over all neighboring datasets. As written, the privacy semantics are ambiguous and weaker than the formal definition used elsewhere in the paper.
minor comments (4)
  1. [Section 3.3] The phrase 'detects local alternatives distinct from the null at the fastest rate of 1/sqrt(n)' is imprecise: the argument shows power tending to 1 when sqrt(K)|mu0_m - mu1_m| -> inf, and only nontrivial power bounded away from alpha when the limit is a finite positive constant. The finite-constant case is not 'detection' in the usual consistency sense and should be stated as such.
  2. [Table 3 and Table 4] In the 'Power 1' columns, the row label (400,200) is repeated three times and the rows (600,300) and (800,400) appear to be missing. This makes the power comparison hard to read and should be corrected.
  3. [Throughout] There are several typographical errors, including 'covariance matrice' in Section 5, 'covairates' in Section 4.2, and inconsistent notation for the tilde/gamma parameters in Remark 3 and Algorithm 1. These should be cleaned up.
  4. [Section 2.2, Remark 1] The statement that any (epsilon+delta,0)-DP algorithm is also (epsilon,delta)-DP is true but the choice of epsilon+delta seems arbitrary and does not by itself justify the focus on Laplace noise; a clearer discussion of the relation between pure and approximate DP would help.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the null distribution and privacy bound are derived from explicit RMT limits and sub-Gaussian tail bounds, not fitted to the data.

full rationale

The paper's central derivations are self-contained and do not reduce to their inputs by construction. Theorem 3 and Lemma 5 derive the asymptotic null distribution of the privatized linear spectral statistics from an explicit CLT: the mean vector and covariance matrix are deterministic integrals of the convolution of the loss functions with the Laplacian noise law against the Marcenko-Pastur limit Fy, with no fitted constants entering the limiting distribution. Theorem 1 and Theorem 2 establish the sensitivity bound analytically from a sub-Gaussian quadratic-form tail bound (Hsu et al., 2012), and the claimed privacy guarantee follows from the Laplace mechanism applied to that bound. The calibration of the noise scale in Algorithm 1 via the preset value gamma_tilde and the private estimate gamma_hat is a practical heuristic; Remark 3 explicitly acknowledges that sigma and gamma are unknown and assumes sigma = 1 and gamma <= gamma_tilde. A legitimate concern—that gamma_hat < gamma with probability tending to 1/2 under H0, making the realized noise scale too small for the stated epsilon-DP guarantee—is a correctness and robustness gap in the two-stage algorithm, not a circularity, because the paper's equations do not define the privacy guarantee in terms of gamma_hat. The self-citations to Zheng et al. (2015, 2019), Yao et al. (2015), and Bai et al. (2009) are used for background, benchmarks, and standard random matrix theory results; they are not load-bearing as the sole justification for the paper's main claims. The asymptotic distribution-free claim rests on the M-P law and the CLT, which are external mathematical results, not on the paper's own fitted values.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The central result rests on modeling assumptions (sub-Gaussian, high-dimensional regime) and regularity conditions on the limiting spectrum; no new physical or statistical entity is invented. The only hand-chosen values are sigma = 1 and gamma_tilde, which are load-bearing for the noise scale.

free parameters (2)
  • preset gamma_tilde = 2 (simulations)
    Used in Algorithm 1 to set the Laplace noise scale. Privacy holds only if the true gamma = Tr(Sigma)/d is not larger than gamma_tilde.
  • sub-Gaussian parameter sigma = 1 (assumed)
    Remark 3 sets sigma = 1 for practical use. If the true parameter exceeds 1, the noise scale is too small and the privacy and null asymptotics can fail.
assumptions (5)
  • domain assumption Population x = Sigma^{1/2} z with z sub-Gaussian, E z = 0, Cov z = I, Tr(Sigma) <= gamma d, ||Sigma|| <= c (Assumption 2).
    Used in Theorem 1 and 2 for the sensitivity bound and throughout for random matrix limit results.
  • domain assumption The dimension-to-sample ratio d/n tends to y in (0, infinity) (Assumption 1).
    Defines the high-dimensional regime needed for the Marcenko-Pastur law and the CLT.
  • ad hoc to paper The limiting spectral support is a finite union of intervals separated and bounded away from 0 (Assumption 4).
    Needed for the eigenvalue rigidity estimates in Lemma 7, but omitted from Theorem 3's stated assumptions; it excludes y = 1.
  • domain assumption The functions g1, g2, g3 satisfy Lyapunov and regularity conditions when evaluated at classical eigenvalue locations plus Laplace noise (Assumptions 5 and 6).
    Verified in the proof of Theorem 3, but stated as general conditions in Lemma 5.
  • standard math External results on eigenvalue rigidity (Knowles-Yin 2017) and a generalized central limit theorem for independent non-identical variables (Shintani-Umeno 2018) apply in the settings used.
    These external theorems are imported without machine-checked proof.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Testing for large-dimensional covariance matrix under differential privacy." pith.science (2026). https://pith.science/paper/OKDOIER5

@misc{pith2026250602410,
  author       = {Pith},
  title        = {Pith review of: Testing for large-dimensional covariance matrix under differential privacy},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OKDOIER5}},
  note         = {Machine review of arXiv:2506.02410}
}
abstract

The increasing prevalence of high-dimensional data across various applications has raised significant privacy concerns in statistical inference. In this paper, we propose a differentially private integrated statistic for testing large-dimensional covariance structures, enabling accurate statistical insights while safeguarding privacy. First, we analyze the global sensitivity of sample eigenvalues for sub-Gaussian populations, where our method bypasses the commonly assumed boundedness of data covariates. For sufficiently large sample size, the privatized statistic guarantees privacy with high probability. Furthermore, when the ratio of dimension to sample size, $d/n \to y \in (0, \infty)$, the privatized test is asymptotically distribution-free with well-known critical values, and detects the local alternative hypotheses distinct from the null at the fastest rate of $1/\sqrt{n}$. Extensive numerical studies on synthetic and real data showcase the validity and powerfulness of our proposed method.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

62 extracted references · 55 canonical work pages

  1. [1]

    Acharya, J., Sun, Z., and Zhang, H. (2018). Differentially private testing of identity and closeness of discrete distributions. In Proceedings of the 32nd International Conference on Neural Information Processing Systems , page 6879–6891

    work page 2018

  2. [2]

    Alabi, D., McMillan, A., Sarathy, J., Smith, A., and Vadhan, S. (2022). Differentially private simple linear regression. Proceedings on Privacy Enhancing Technologies , 2022(2):184--204

    work page 2022

  3. [3]

    and Vadhan, S

    Alabi, D. and Vadhan, S. (2022). Hypothesis testing for differentially private linear regression. In Proceedings of the 36th International Conference on Neural Information Processing Systems

    work page 2022

  4. [4]

    Amin, K., Dick, T., Kulesza, A., Medina, A. M. n., and Vassilvitskii, S. (2019). Differentially private covariance estimation. In Proceedings of the 33rd International Conference on Neural Information Processing Systems , pages 14213--14222

    work page 2019

  5. [5]

    Anderson, T. W. (2003). An introduction to multivariate statistical analysis . Wiley New York

    work page 2003

  6. [6]

    Bai, Z., Jiang, D., Yao, J., and Zheng, S. (2009). Corrections to LRT on large-dimensional covariance matrix by RMT . The Annals of Statistics , 37(6):3822--3840

    work page 2009

  7. [7]

    and Silverstein, J

    Bai, Z. and Silverstein, J. W. (2004). CLT for linear spectral statistics of large-dimensional sample covariance matrices. The Annals of Probability , 32(1A):553--605

    2004

  8. [8]

    and Dette, H

    Birke, M. and Dette, H. (2005). A note on testing the covariance matrix for large dimension. Statistics & Probability Letters , 74(3):281--289

    work page 2005

Show all 62 references
  1. [9]

    Blum, A., Dwork, C., McSherry, F., and Nissim, K. (2005). Practical privacy: the SuLQ framework. In Proceedings of the Twenty-Fourth ACM SIGMOD-SIGACT-SIGART Symposium on Principles of Database Systems , page 128–138

    2005

  2. [10]

    Cai, T. T. and Ma, Z. (2013). Optimal hypothesis testing for high dimensional covariance matrices. Bernoulli , 19(5B):2359--2388

    2013

  3. [11]

    T., Wang, Y., and Zhang, L

    Cai, T. T., Wang, Y., and Zhang, L. (2021). The cost of privacy: Optimal rates of convergence for parameter estimation with differential privacy. The Annals of Statistics , 49(5):2825--2850

    2021

  4. [12]

    T., Xia, D., and Zha, M

    Cai, T. T., Xia, D., and Zha, M. (2024). Optimal differentially private PCA and estimation for spiked covariance matrices. arXiv preprint arXiv:2401.03820

    2024 arXiv

  5. [13]

    Canonne, C., Kamath, G., McMillan, A., Ullman, J., and Zakynthinou, L. (2020). Private identity testing for high-dimensional distributions. In Proceedings of the 34th International Conference on Neural Information Processing Systems , pages 10099 -- 10111

    2020

  6. [14]

    L., Kamath, G., McMillan, A., Smith, A., and Ullman, J

    Canonne, C. L., Kamath, G., McMillan, A., Smith, A., and Ullman, J. (2019). The structure of optimal private tests for simple hypotheses. In Proceedings of the 51st Annual ACM SIGACT Symposium on Theory of Computing , pages 310--321

    2019

  7. [15]

    Chang, J., He, J., Kang, J., and Wu, M. (2024). Statistical inferences for complex dependence of multimodal imaging data. Journal of the American Statistical Association , 119(546):1486--1499

    2024

  8. [16]

    Chaudhuri, K., Monteleoni, C., and Sarwate, A. D. (2011). Differentially private empirical risk minimization. Journal of Machine Learning Research , 12(3):1069--1109

    2011

  9. [17]

    Couch, S., Kazan, Z., Shi, K., Bray, A., and Groce, A. (2019). Differentially private nonparametric hypothesis testing. In Proceedings of the 2019 ACM SIGSAC Conference on Computer and Communications Security , pages 737--751

    2019

  10. [18]

    Ding, B., Kulkarni, J., and Yekhanin, S. (2017). Collecting telemetry data privately. In Proceedings of the 31st International Conference on Neural Information Processing Systems , page 3574–3583

    2017

  11. [19]

    Ding, B., Nori, H., Li, P., and Allen, J. (2018). Comparing population means under local differential privacy: with significance and power. In Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence and Thirtieth Innovative Applications of Artificial Intell...

    2018

  12. [20]

    B., and Butucea, C

    Dubois, A., Berrett, T. B., and Butucea, C. (2019). Goodness-of-fit testing for H \"o lder continuous densities under local differential privacy. In Foundations of Modern Statistics , pages 53--119. Springer

    2019

  13. [21]

    Dunsche, M., Kutta, T., and Dette, H. (2022). Multivariate mean comparison under differential privacy. In International Conference on Privacy in Statistical Databases , pages 31--45

    2022

  14. [22]

    Dwork, C., Kenthapadi, K., McSherry, F., Mironov, I., and Naor, M. (2006a). Our data, ourselves: Privacy via distributed noise generation. In Advances in Cryptology-EUROCRYPT 2006 , pages 486--503

    2006

  15. [23]

    Dwork, C., McSherry, F., Nissim, K., and Smith, A. (2006b). Calibrating noise to sensitivity in private data analysis. In Theory of Cryptography , pages 265--284

    2006

  16. [24]

    Dwork, C., Roth, A., et al. (2014a). The algorithmic foundations of differential privacy. Foundations and Trends in Theoretical Computer Science , 9(3--4):211--407

    2014

  17. [25]

    Dwork, C., Talwar, K., Thakurta, A., and Zhang, L. (2014b). Analyze Gauss : optimal bounds for privacy-preserving principal component analysis. In Proceedings of the Forty-Sixth Annual ACM Symposium on Theory of Computing , pages 11--20

    2014

  18. [26]

    Erlingsson, \'U ., Pihur, V., and Korolova, A. (2014). Rappor: Randomized aggregatable privacy-preserving ordinal response. In Proceedings of the 2014 ACM SIGSAC Conference on Computer and Communications Security , pages 1054--1067

    2014

  19. [27]

    Fan, J., Han, F., and Liu, H. (2014). Challenges of big data analysis. National Science Review , 1(2):293--314

    2014

  20. [28]

    Gaboardi, M., Lim, H., Rogers, R., and Vadhan, S. (2016). Differentially private chi-squared hypothesis testing: Goodness of fit and independence testing. In Proceedings of The 33rd International Conference on Machine Learning , pages 2111--2120

    2016

  21. [29]

    Hamilton, J. D. (2020). Time series analysis . Princeton university press

    2020

  22. [30]

    H \"a rdle, W. K. and Simar, L. (2019). Applied multivariate statistical analysis . Springer

    2019

  23. [31]

    Hsiao, C. (2022). Analysis of panel data . Cambridge university press

    2022

  24. [32]

    Hsu, D., Kakade, S., and Zhang, T. (2012). A tail inequality for quadratic forms of subgaussian random vectors . Electronic Communications in Probability , 17:1--6

    2012

  25. [33]

    Huang, C., Wang, D., Hu, Y., and Sartori, N. (2022). Comparison of two mean vectors under differential privacy for high-dimensional data. In Proceedings of the 4th International Conference on Statistics: Theory and Applications

    2022

  26. [34]

    Jiang, D., Jiang, T., and Yang, F. (2012). Likelihood ratio tests for covariance matrices of high-dimensional normal distributions. Journal of Statistical Planning and Inference , 142(8):2241--2256

    2012

  27. [35]

    Johnstone, I. M. (2001). On the distribution of the largest eigenvalue in principal components analysis. The Annals of Statistics , 29(2):295--327

    2001

  28. [36]

    and Schrab, A

    Kim, I. and Schrab, A. (2023). Differentially private permutation tests: Applications to kernel methods. arXiv preprint arXiv:2310.19043

    2023 arXiv

  29. [37]

    and Yin, J

    Knowles, A. and Yin, J. (2017). Anisotropic local laws for random matrices. Probability Theory and Related Fields , 169:257--352

    2017

  30. [38]

    Lam Weil, J., Laurent, B., and Loubes, J. M. (2022). Minimax optimal goodness-of-fit testing for densities and multinomials under a local differential privacy constraint. Bernoulli , 28(1):579--600

    2022

  31. [39]

    and Wolf, M

    Ledoit, O. and Wolf, M. (2002). Some hypothesis tests for the covariance matrix when the dimension is large compared to the sample size. The Annals of Statistics , 30(4):1081--1102

    2002

  32. [40]

    L., Romano, J

    Lehmann, E. L., Romano, J. P., and Casella, G. (2008). Testing statistical hypotheses . Springer

    2008

  33. [41]

    Liu, X., Kong, W., Jain, P., and Oh, S. (2024). DP-PCA : statistically optimal and differentially private PCA . In Proceedings of the 36th International Conference on Neural Information Processing Systems , pages 29929--29943

    2024

  34. [42]

    Marchenko, V. A. and Pastur, L. A. (1967). Distribution of eigenvalues for some sets of random matrices. Matematicheskii Sbornik , 114(4):507--536

    1967

  35. [43]

    A., Vannatta, R

    Mertler, C. A., Vannatta, R. A., and LaVenia, K. N. (2021). Advanced and multivariate statistical methods: Practical application and interpretation . Routledge

    2021

  36. [44]

    Narayanan, S. (2022). Private high-dimensional hypothesis testing. In Proceedings of the 35th Annual Conference on Learning Theory , pages 3979--4027. PMLR

    2022

  37. [45]

    P \'e ch \'e , S. (2009). Universality results for the largest eigenvalues of some sample covariance matrix ensembles. Probability Theory and Related Fields , 143(3):481--516

    2009

  38. [46]

    and Chen, S

    Qiu, Y. and Chen, S. (2015). Bandwidth selection for high-dimensional covariance matrix estimation. Journal of the American Statistical Association , 110(511):1160--1174

    2015

  39. [47]

    Raj, A., Law, H. C. L., Sejdinovic, D., and Park, M. (2019). A differentially private kernel two-sample test. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases , pages 697--724

    2019

  40. [48]

    Sheffet, O. (2017). Differentially private ordinary least squares. In Proceedings of the 34th International Conference on Machine Learning , page 3105–3114

    2017

  41. [49]

    and Umeno, K

    Shintani, M. and Umeno, K. (2018). Super generalized central limit theorem—limit distributions for sums of non-identical random variables with power laws. Journal of the Physical Society of Japan , 87(4):043003

    2018

  42. [50]

    Soshnikov, A. (2002). A note on universality of the distribution of the largest eigenvalues in certain sample covariance matrices. Journal of Statistical Physics , 108:1033--1056

    2002

  43. [51]

    Srivastava, M. S. (2005). Some tests concerning the covariance matrix in high dimensional data. Journal of the Japan Statistical Society , 35(2):251--272

    2005

  44. [52]

    Srivastava, M. S. and Reid, N. (2012). Testing the structure of the covariance matrix with fewer observations than the dimension. Journal of Multivariate Analysis , 112:156--171

    2012

  45. [53]

    Taherpour, A., Taherpour, A., Khattab, T., and Wang, X. (2024). Large array antenna spectrum sensing in cognitive radio networks. IEEE Transactions on Cognitive Communications and Networking

    2024

  46. [54]

    Team, A. D. P. (2017). Learning with privacy at scale. Apple Machine Learning Journal , 1(8):1--25

    2017

  47. [55]

    Wang, C., Yang, J., Miao, B., and Cao, L. (2013). Identity tests for high dimensional data using RMT . Journal of Multivariate Analysis , 118:128--137

    2013

  48. [56]

    Wang, Y. (2018). Revisiting differentially private linear regression: optimal and adaptive prediction & estimation in unbounded domain. arXiv preprint arXiv:1803.02596

    2018 arXiv

  49. [57]

    Wang, Y., Lee, J., and Kifer, D. (2015). Revisiting differentially private hypothesis tests for categorical data. arXiv preprint arXiv:1511.03376

    2015 arXiv

  50. [58]

    Yao, J., Zheng, S., and Bai, Z. (2015). Large sample covariance matrices and high-dimensional data analysis . Cambridge University Press

    2015

  51. [59]

    and Zou, H

    Yi, F. and Zou, H. (2013). Sure-tuned tapering estimation of large covariance matrices. Computational Statistics & Data Analysis , 58:339--351

    2013

  52. [60]

    Zheng, S., Bai, Z., and Yao, J. (2015). Substitution principle for CLT of linear spectral statistics of high-dimensional sample covariance matrices with applications to hypothesis testing. The Annals of Statistics , 43(2):546--591

    2015

  53. [61]

    Zheng, S., Chen, Z., Cui, H., and Li, R. (2019). Hypothesis testing on linear structures of high dimensional covariance matrix. The Annals of statistics , 47(6):3300 -- 3334

    2019

  54. [62]

    Zhu, X., Guo, J., Guo, X., Zhu, L., and Zheng, J. (2023). Bandwidth selection for large covariance and precision matrices. Statistica Sinica , pages 321--341

    2023

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.