Pith. sign in

REVIEW 3 major objections 5 minor 28 references

Optimizing Noise Distributions for Differential Privacy

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read Noise distributions optimized for a fixed composition count and variance deliver smaller (ε, δ)-DP guarantees than Gaussian, Laplace, Staircase, or Cactus noise in the moderate composition regime.

desk verdict A credible, reproducible empirical improvement over Gaussian/Laplace noise for moderate composition, with the 'optimal' label running ahead of the actual guarantees. read the letter →

arxiv 2504.14730 v2 pith:QWM3OJOM submitted 2025-04-20 cs.IT math.IT

Atefeh Gilani , Juan Felipe Gomez , Shahab Asoodeh , Flavio P. Calmon , Oliver Kosut , Lalitha Sankar This is my paper · ORCID
classification cs.ITmath.IT MSC 68P2794A17
keywords differentialprivacyRenyinoisedistributionoptimizationmomentsaccountantconvexpreconditionedgradientdescentmoderatecompositionregimediscreteGaussian
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 claims that choosing the noise distribution to match the exact number of times a query will be composed gives better differential privacy than the standard Gaussian or Laplace mechanisms at the same noise variance. The optimization minimizes Renyi differential privacy for an order $\alpha$ that is itself tuned by the moments accountant, which lets one framework interpolate between the single-composition regime, where it recovers the Staircase distribution, and the large-composition regime, where it recovers the Cactus distribution. In the moderate regime of roughly 10 to 40 compositions, the optimized continuous and discrete distributions achieve a smaller $\varepsilon$ at the same $\delta$ and variance—about 8% to 12% lower mean-squared error on real datasets when the privacy budget is fixed. Real deployments such as repeated census releases or machine-learning training with a limited number of rounds sit precisely in this regime, so the improvement applies where standard mechanisms are routinely used.

What carries the argument

The load-bearing object is $g_\alpha(P_Z) = \max_{t\in S} \mathbb{E}_{P_Z}\left[\left(P_Z/T_t P_Z\right)^{\alpha-1}\right]$, the worst-case term inside the logarithm of the Renyi divergence of order $\alpha$ over shifts $t$ within the sensitivity set $S$. Theorem 3.1 proves that $g_\alpha(P_Z)$ is convex in the distribution and that symmetry is sufficient, so after restricting to symmetric piecewise-constant densities (or PMFs) with geometric tails—Definitions 3.3 and 3.4—the search becomes a finite-dimensional convex problem over bin probabilities $p_0,\dots,p_N$ (Theorem 3.6). The solver is a preconditioned gradient descent with scaling matrix $M=\mathrm{diag}(p)^{-1}$, which keeps iterates strictly inside the positive orthant and thereby avoids the infinite Renyi divergence that arises when supports shrink, and a Newton update of $\alpha$ that follows the moments-accountant objective $\gamma_{p,N_c,\delta}(\alpha)$.

What would settle it

For $s=1$, $\sigma=8$, $N_c=10$, $\delta=10^{-6}$, search a much finer family—a high-resolution histogram with no geometric-tail restriction and a dense grid over $\alpha$—and evaluate every candidate with Connect-the-Dots; any distribution with variance $64$ and $\varepsilon$ below $1.62$ would refute the paper's optimality claim.

Watch

Extended reading notes

Core claim

For a fixed sensitivity $s$, noise variance $\sigma^2$, target $\delta$, and composition count $N_c$, the paper constructs a symmetric, piecewise-constant density (or, for integer outputs, a symmetric probability mass function) with geometric tails that minimizes the Renyi differential privacy parameter at an order $\alpha$ selected to track the moments-accountant optimum for $N_c$ and $\delta$. The central discovery is that this distribution dominates Gaussian, Laplace, Staircase, and Cactus noise at the same variance in the moderate composition regime (roughly 10 to 40 compositions), with the largest gaps appearing for moderate $\alpha$ values, and that the same framework automatically reproduces Cactus for very large $N_c$ and Staircase for a single composition. All reported $(\varepsilon,\delta)$ curves are evaluated with Connect-the-Dots accounting, so the comparison uses the same tight privacy accountant practitioners rely on.

Load-bearing premise

The paper assumes that the best noise distribution is well captured by a symmetric step-shaped density whose tails decay geometrically from a hand-chosen bin, and that a Newton shortcut finds the right Renyi order $\alpha$; if the true optimum lives outside this family or the shortcut stalls at a local optimum, the output is only a local optimum, not the global one.

Editorial extensions

If this is right

  • For any deployment with a known composition count in the moderate regime, the optimized mechanism gives a smaller $\varepsilon$ than Gaussian or Laplace at the same variance, which is a privacy improvement at zero added noise.
  • Practitioners no longer need to choose among Staircase, Cactus, Gaussian, and Laplace: the same algorithm selects the right shape automatically as $\alpha$ varies with $N_c$ and $\delta$.
  • For integer-valued queries, the discrete version yields noise that beats discrete Gaussian and discrete Laplace noise while staying immune to floating-point attacks.
  • Because the optimized Renyi DP depends only on the ratio $\sigma/s$, the same relative noise level transfers across query sensitivities.
  • The reported gains translate into roughly 8 to 12% lower mean-squared error on real datasets when the privacy budget is held fixed.

Reading between the lines

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

  • A testable extension is to relax the geometric-tail restriction: comparing against a much larger $N$ and finer bin width, or a non-geometric tail, would quantify whether the family restriction loses anything in the moderate regime.
  • The same recipe—optimize a Renyi surrogate at a tuned $\alpha$, then convert via the moments accountant—could be applied to other cost functions such as $\ell^1$ cost or heterogeneous per-query costs, where symmetric variance-constrained shapes may not be optimal.
  • A practical consequence the paper does not develop: for machine-learning training with a moderate number of update rounds, replacing Gaussian noise with this mechanism could lower the noise multiplier at a fixed privacy budget, provided adaptive composition and subsampling behave for this noise as they do for Gaussian.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes a unified optimization framework for designing continuous and discrete noise distributions for differential privacy by minimizing Rényi divergence of a tunable order α under a cost (variance) constraint. The infinite-dimensional distribution search is reduced to a finite-dimensional convex problem over symmetric piecewise-constant (continuous) or PMF (discrete) distributions with geometric tails, controlled by hyperparameters N, r, and Δ. The resulting problem is solved by a preconditioned gradient descent algorithm that also updates α heuristically via a Newton step based on the moments accountant. The authors demonstrate that the method recovers the Staircase and Cactus mechanisms in the single- and large-composition limits, and report improved ε for fixed δ and σ in moderate composition regimes, with all final privacy guarantees computed using the Connect-the-Dots accountant.

Significance. If the reported gains are stable, the paper delivers a practical and unified recipe for noise design that outperforms the standard Gaussian and Laplace mechanisms at equal variance in moderate composition regimes, and the recovery of known optimal mechanisms in limiting regimes is a strong sanity check. The paper also ships code, evaluates with a state-of-the-art accountant, and provides a clean convexity reduction (Theorem 3.1) that is rigorous. The main caveats are that the 'optimal' label is tied to a surrogate moments-accountant objective and to a hand-chosen parametric family, and the numerical optimization lacks convergence guarantees; these issues do not invalidate the empirical comparison but do narrow the scope of the claims.

major comments (3)
  1. [§3.1, Definitions 3.3–3.4; §5] The finite-dimensional restriction to symmetric piecewise-constant distributions with geometric tails is asserted (Section 3.1: 'this tail assumption does not significantly impact the optimization, provided that the interval is sufficiently large') to closely approximate the true optimizer, but no approximation error bound or guidance on N, r, and Δ is given. The numerical sections fix values such as N=8000, Δ=0.01, r=0.9999 with no sensitivity analysis. If the reported ε at the target δ changes materially when N is doubled or Δ is halved, the headline improvements in Figures 4–5 and Table 1 are artifacts of under-optimization rather than properties of the true optimizer within the family. Please add a sensitivity study or prove an approximation bound before claiming optimality.
  2. [§4, Algorithm 3; Algorithm 1] The optimization minimizes the moments-accountant surrogate inf_α [N_c/(α−1) log g_α(p) + log(1/δ)/(α−1)] (Eq. (25)), not the true ε or the Connect-the-Dots accountant used for evaluation. Algorithm 1's output description ('Optimal distribution P* minimizing ε') is therefore overstated. Moreover, the preconditioned gradient descent on the nonsmooth objective max_t g_α(p,t) with a backtracking line search over a finite geometric step-size sequence, together with the Newton update for α (line 21), has no convergence or global-optimality guarantee. The statement in Section 4 that 'gradient descent will converge to the optimal objective value' is not justified. Please either provide a convergence analysis or rephrase to 'heuristic optimization' and add empirical evidence such as multiple restarts or a comparison with a certified convex solver for fixed α.
  3. [Theorem 3.6, Eq. (23); Appendix C] The reduction of the continuous optimization to shifts t in {1,...,s/Δ} relies on piecewise linearity of g_α in t. The derivation in Appendix C is plausible, but the final closed form in Eq. (23) is dense and I could not independently verify every index; for instance, the last sum mixes p^α_{|t+j|} with r^{-j(1−α)} and the correspondence with Eq. (75) of Appendix C is not immediate. Since (23) is exactly the objective that the convex solver minimizes, please provide a cleaner step-by-step derivation, a machine-checkable verification, or a statement that the algebra has been symbolically verified, and explicitly specify how the case a=0 is handled in the sums.
minor comments (5)
  1. [Algorithm 3, line 10] The definition of μ_ub assumes that at least one component of g_proj is positive; if all components are non-positive, μ_ub is undefined. Please specify the fallback behavior (e.g., terminate the line search or set μ_ub to a large value).
  2. [Algorithm 3, line 21] The Newton update for α is unconstrained and may push α below 1, which is outside the valid range for Rényi divergence. Please add a safeguard such as projecting α to [1+ε, ∞).
  3. [Figure 4 caption and surrounding text] The caption says RDP noise is marked as the winner when it achieves more than a 2% improvement, but the text then says 'Even when another distribution is marked as the best, our noise still consistently outperforms the others, although by less than 2%.' This is confusing; please clarify the exact criterion for marking the winner and what is meant by 'outperforms' in the latter sentence.
  4. [Table 1 and Appendix D] The text states that Gaussian noise consistently outperformed Laplace, Cactus, and Staircase across all tested settings, but the table only reports improvements over Gaussian. Since the paper's contribution is comparison with all baselines, please include the relative performance against the other baselines for at least one representative setting, or provide a supplementary table.
  5. [Eq. (25) and Algorithm 3] The notation γ' and γ'' for the Newton update refers to derivatives with respect to α, but γ_{p,N_c,δ}(α) depends on α both through the explicit denominator and through g_α(p). Please clarify whether the derivatives account for the α-dependence of g_α(p) and how the derivative is computed numerically.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the RDP/moments-accountant surrogate used for optimization is distinct from the Connect-the-Dots accountant used for evaluation, so the reported gains are empirical rather than true by construction.

full rationale

The paper's core empirical claim is that its optimized RDP noise achieves smaller (epsilon, delta) at fixed variance in moderate composition regimes. The optimization (Algorithms 1-3, Eq. 22, Eq. 25) minimizes a finite-dimensional convex surrogate based on Renyi divergence and the simple moments-accountant bound of Proposition 2.4. Section 5 explicitly states that 'All (epsilon, delta)-DP guarantees presented here are computed using the Connect-the-Dots accounting (Doroshenko et al., 2022)', a different, external accountant. Hence the reported epsilon values are not the optimized objective, and the superiority over Gaussian and Laplace is not an identity or a fitted tautology. Recovery of Staircase (alpha toward infinity) and Cactus (alpha toward 1) is presented as a numerical sanity check against independently known mechanisms, not as the source of the moderate-regime claim; no uniqueness theorem from the authors' prior work is invoked to exclude alternatives. The finite-dimensional piecewise-constant geometric-tail family (Definitions 3.3-3.4) is a computational restriction, and Section 3.1 concedes without proof that the tail restriction 'does not significantly impact the optimization, provided that the interval is sufficiently large.' That is a missing approximation bound, not a circular reduction; likewise, the Newton alpha update in Algorithm 3 has no convergence guarantee, and the hyperparameters N, r, delta are not subjected to sensitivity analysis. These are robustness and correctness concerns, not circularity. Self-citations to Alghamdi et al. (2022, 2023) and Asoodeh et al. (2021) provide baselines and context, but the central comparison is self-contained against external benchmarks, so the circularity score is 0.

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

The central claim rests on standard DP analysis and a hand-chosen finite-dimensional family. The free parameters N, r, and Delta are not fitted to external data, but they shape the optimization and are set per experiment. No new entities are postulated, and the paper introduces no new forces, particles, or conserved quantities.

free parameters (4)
  • N (number of bins in distribution family) = 1600 to 8000 in experiments
    Hyperparameter determining the number of probability bins before geometric tails; chosen by hand per setting (Figure 2 and Figure 3 captions).
  • r (geometric tail decay rate) = 0.9999 in experiments
    Controls tail decay; set to 0.9999 in all experiments, affecting the tail shape and hence the optimized distribution.
  • Delta (bin width) = 0.005 to 0.01 in experiments, subject to s/Delta integer
    Bin width for piecewise constant densities; chosen by hand, constrains the variance floor and the support resolution.
  • T (alpha update time step) and learning rate schedule = not specified precisely
    Algorithm hyperparameters that affect convergence but not the final solution if converged; still chosen by hand.
assumptions (6)
  • standard math g_alpha(P) = max_t E_P[(P / T_t P)^(alpha-1)] is convex in P for alpha > 1
    Invoked in Theorem 3.1, citing van Erven and Harremos; used to justify the convex finite-dimensional optimization.
  • standard math A symmetric distribution suffices for the optimal noise
    Proven in Appendix A via convexity and symmetry of the feasible region.
  • domain assumption The worst-case shift for a piecewise constant PDF occurs at integer multiples of the bin width
    Used in Theorem 3.6 to reduce the continuous maximization over t in [-s, s] to a finite set; argued via piecewise linearity in Appendix C.
  • ad hoc to paper The geometric-tail family (Definitions 3.3 and 3.4) approximates the true optimum well
    Assumption in Section 3.1 that restricting to this finite-dimensional family does not significantly affect the optimum; no error bound provided.
  • domain assumption Moments accountant conversion epsilon = gamma(alpha) + log(1/delta) / (alpha-1) with optimal alpha
    Used as the optimization objective (Section 4, equation 25); standard result from Abadi et al. and Mironov.
  • domain assumption Connect-the-Dots accounting correctly certifies the (epsilon, delta) guarantees of the optimized distributions
    Relied upon for all numerical privacy curves (Section 5); this is a state-of-the-art accountant from Doroshenko et al.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Optimizing Noise Distributions for Differential Privacy." pith.science (2026). https://pith.science/paper/QWM3OJOM

@misc{pith2026250414730,
  author       = {Pith},
  title        = {Pith review of: Optimizing Noise Distributions for Differential Privacy},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QWM3OJOM}},
  note         = {Machine review of arXiv:2504.14730}
}
abstract

We propose a unified optimization framework for designing continuous and discrete noise distributions that ensure differential privacy (DP) by minimizing R\'enyi DP, a variant of DP, under a cost constraint. R\'enyi DP has the advantage that by considering different values of the R\'enyi parameter $\alpha$, we can tailor our optimization for any number of compositions. To solve the optimization problem, we reduce it to a finite-dimensional convex formulation and perform preconditioned gradient descent. The resulting noise distributions are then compared to their Gaussian and Laplace counterparts. Numerical results demonstrate that our optimized distributions are consistently better, with significant improvements in $(\varepsilon, \delta)$-DP guarantees in the moderate composition regimes, compared to Gaussian and Laplace distributions with the same variance.

Figures

Figures reproduced from arXiv: 2504.14730 by the authors.

Figure 1
Figure 1. This plot compares the privacy curve of our opti￾mized noise—designed for 10 compositions, s = 1, and δ = 10−6—against other noise distributions, all with the same standard deviation of 8. At the target δ = 10−6 , our noise achieves an ε of 1.62, compared to 1.76 for Laplace/Staircase and 1.74 for Gaussian/Cactus. Replacing these mechanisms with our optimized noise results in at least a 6.89% improvement in the ε va… view at source ↗
Figure 2
Figure 2. The left plot compares our optimized noise distribution and a Cactus distribution, both with standard deviation 0.3, under the setting δ = 10−5 , s = 1, and 50,000 compositions, corresponding to α = 1.006 (moments accountant), with noise parameters ∆ = 0.005, N = 1600, and r = 0.9999. The right plot compares our optimized noise and a Staircase distribution, both with standard deviation 1, under the setting δ = 10−20… view at source ↗
Figure 3
Figure 3. The left plot displays the optimized noise distribution for a standard deviation of 4, δ = 10−6 , s = 1, and 20 compositions, corresponding to α = 5.95 (using the moments accountant). The right plot shows the distribution for a standard deviation of 2, δ = 10−10 , s = 1, and 8 compositions, corresponding to α = 11.32. The noise parameters are ∆ = 0.01, r = 0.9999, and N = 8000 for the left plot, and N = 4000 for the… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: For fixed δ = 10−6 , the grid evaluates different combinations of composition count and noise standard deviation (σ), marking RDP noise as the winner whenever it achieves more than a 2% improvement in ε over the best alternative (among Gaussian, Laplace, Staircase, and…
Figure 5
Figure 5. Figure 5: The plots compare the pointwise minimum of privacy curves from our optimized noise distributions—designed for 10 composi￾tions, sensitivity 1, and varying δ—with privacy curves for Gaussian, Laplace, Cactus, and Staircase mechanisms, all with a fixed standard deviation…
Figure 6
Figure 6. Figure 6: For fixed δ = 10−9 , the grid evaluates different combinations of composition count and noise standard deviation (σ), marking RDP noise as the winner whenever it achieves more than a 2% improvement in ε over the best alternative (among Gaussian, Laplace, Staircase, and…
Figure 7
Figure 7. Figure 7: The plots compare the pointwise minimum of privacy curves from our optimized noise distributions—designed for 10 composi￾tions, sensitivity 1, and varying δ—with privacy curves for Gaussian, Laplace, Cactus, and Staircase mechanisms, all with a fixed standard deviation…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 16 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    https://github.com/SankarLab/Renyi-DP-Mechanism-Design, May 2025

    Renyi DP mechanism design. https://github.com/SankarLab/Renyi-DP-Mechanism-Design, May 2025

    work page 2025

  3. [3]

    B., Mironov, I., Talwar, K., and Zhang, L

    Abadi, M., Chu, A., Goodfellow, I., McMahan, H. B., Mironov, I., Talwar, K., and Zhang, L. Deep learning with differential privacy. In Proceedings of the 2016 ACM SIGSAC conference on computer and communications security, pp.\ 308--318, 2016

    2016

  4. [4]

    The 2020 Census Disclosure Avoidance System TopDown Algorithm

    Abowd, J., Ashmead, R., Cumings-Menon, R., Garfinkel, S., Heineck, M., Heiss, C., Johns, R., Kifer, D., Leclerc, P., Machanavajjhala, A., Moran, B., Sexton, W., Spence, M., and Zhuravlev, P. The 2020 Census Disclosure Avoidance System TopDown Algorithm . Harvard Data Science Review, 0 (Special Issue 2), jun 24 2022. https://hdsr.mitpress.mit.edu/pub/7evz361i

    work page 2020

  5. [5]

    P., Kosut, O., Sankar, L., and Wei, F

    Alghamdi, W., Asoodeh, S., Calmon, F. P., Kosut, O., Sankar, L., and Wei, F. Cactus mechanisms: Optimal differential privacy mechanisms in the large-composition regime. In 2022 IEEE International Symposium on Information Theory (ISIT), pp.\ 1838--1843, 2022. doi:10.1109/ISIT50566.2022.9834438

    arXiv 2022

  6. [6]

    P., Felipe Gomez, J., Kosut, O., and Sankar, L

    Alghamdi, W., Asoodeh, S., Calmon, F. P., Felipe Gomez, J., Kosut, O., and Sankar, L. Schrödinger mechanisms: Optimal differential privacy mechanisms for small sensitivity. In 2023 IEEE International Symposium on Information Theory (ISIT), pp.\ 2201--2206, 2023. doi:10.1109/ISIT54713.2023.10206616

    arXiv 2023

  7. [7]

    P., Kosut, O., and Sankar, L

    Asoodeh, S., Liao, J., Calmon, F. P., Kosut, O., and Sankar, L. Three variants of differential privacy: Lossless conversion and applications. IEEE Journal on Selected Areas in Information Theory, 2 0 (1): 0 208--222, 2021

    work page 2021

  8. [8]

    and Vadhan, S

    Balcer, V. and Vadhan, S. P. Differential privacy on finite computers. In Information Technology Convergence and Services, 2017. URL https://api.semanticscholar.org/CorpusID:6692720

    work page 2017

Show all 28 references
  1. [9]

    Hypothesis testing interpretations and R\'enyi differential privacy

    Balle, B., Barthe, G., Gaboardi, M., Hsu, J., and Sato, T. Hypothesis testing interpretations and R\'enyi differential privacy. In Int. Conf. Art. Intelligence and Stat. (AISTAT), pp.\ 2496--2506, 2020

    2020

  2. [10]

    and Steinke, T

    Bun, M. and Steinke, T. Concentrated differential privacy: Simplifications, extensions, and lower bounds. In Proc. Theory of Cryptography, pp.\ 635--658, 2016

    2016

  3. [11]

    L., Kamath, G., and Steinke, T

    Canonne, C. L., Kamath, G., and Steinke, T. The discrete Gaussian for differential privacy. In Proceedings of the 34th International Conference on Neural Information Processing Systems, NIPS '20, Red Hook, NY, USA, 2020. Curran Associates Inc. ISBN 9781713829546

    2020

  4. [12]

    G., and Woodruff, D

    Chen, Y., Thakurta, A. G., and Woodruff, D. P. Poisson subsampled rényi differential privacy. In Proceedings of the 52nd Annual ACM Symposium on Theory of Computing (STOC), pp.\ 849--860. ACM, 2019. doi:10.1145/3313276.3316369

    2019

  5. [13]

    Connect the dots: Tighter discrete approximations of privacy loss distributions, 2022

    Doroshenko, V., Ghazi, B., Kamath, P., Kumar, R., and Manurangsi, P. Connect the dots: Tighter discrete approximations of privacy loss distributions, 2022. URL https://arxiv.org/abs/2207.04380

    2022 arXiv

  6. [14]

    Our data, ourselves: Privacy via distributed noise generation

    Dwork, C., Kenthapadi, K., McSherry, F., Mironov, I., and Naor, M. Our data, ourselves: Privacy via distributed noise generation. In Vaudenay, S. (ed.), EUROCRYPT, pp.\ 486--503, 2006 a

    2006

  7. [15]

    Calibrating noise to sensitivity in private data analysis

    Dwork, C., McSherry, F., Nissim, K., and Smith, A. Calibrating noise to sensitivity in private data analysis. In Proc. Theory of Cryptography (TCC), pp.\ 265--284, Berlin, Heidelberg, 2006 b . ISBN 3-540-32731-2, 978-3-540-32731-8

    2006

  8. [16]

    N., and Rothblum, G

    Feldman, V., Acharya, A., Dinesh, P. N., and Rothblum, G. N. On the rényi differential privacy of the shuffle model. In Proceedings of the 52nd Annual ACM Symposium on Theory of Computing (STOC), pp.\ 1232--1244. ACM, 2021. doi:10.1145/3460120.3484794

    2021

  9. [17]

    V., Muthukrishnan, S., and Kuczynski, J

    Feldman, V., Liu, Y., Karpov, D. V., Muthukrishnan, S., and Kuczynski, J. Stronger privacy amplification by shuffling for rényi and approximate differential privacy. arXiv preprint arXiv:2208.04591, 2022. URL https://arxiv.org/abs/2208.04591

    2022 arXiv

  10. [18]

    The staircase mechanism in differential privacy

    Geng, Q., Kairouz, P., Oh, S., and Viswanath, P. The staircase mechanism in differential privacy. IEEE Journal of Selected Topics in Signal Processing, 9 0 (7): 0 1176--1184, 2015. doi:10.1109/JSTSP.2015.2425831

    2015

  11. [19]

    Universally utility-maximizing privacy mechanisms

    Ghosh, A., Roughgarden, T., and Sundararajan, M. Universally utility-maximizing privacy mechanisms. SIAM J. Comput., 41 0 (6): 0 1673–1693, January 2012. ISSN 0097-5397. doi:10.1137/09076828X. URL https://doi.org/10.1137/09076828X

    2012 doi

  12. [20]

    Heart disease data set

    Janosi, A., Steinbrunn, W., Pfisterer, M., and Detrano, R. Heart disease data set. https://archive.ics.uci.edu/ml/datasets/Heart+Disease, 1988. Accessed: 2025-04-29

    1988

  13. [21]

    Diabetes dataset

    learn developers, S. Diabetes dataset. https://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_diabetes.html. Accessed: 2025-04-29

    2025

  14. [22]

    M., Smith, A., and Vadhan, S

    Lécuyer, M., Lee, J. M., Smith, A., and Vadhan, S. Practical privacy filters and odometers with rényi differential privacy and applications to differentially private deep learning. arXiv preprint arXiv:2103.01379, 2021. URL https://arxiv.org/abs/2103.01379

    2021 arXiv

  15. [23]

    On significance of the least significant bits for differential privacy

    Mironov, I. On significance of the least significant bits for differential privacy. In Proceedings of the 2012 ACM Conference on Computer and Communications Security, CCS '12, pp.\ 650–661, New York, NY, USA, 2012. Association for Computing Machinery. ISBN 9781450316514. doi:1...

    2012

  16. [24]

    R\'enyi differential privacy

    Mironov, I. R\'enyi differential privacy. In Proc. IEEE Comp. Security Foundations Symp. (CSF), pp.\ 263--275, 2017

    2017

  17. [25]

    and Harremos, P

    van Erven, T. and Harremos, P. Rényi divergence and kullback-leibler divergence. IEEE Transactions on Information Theory, 60 0 (7): 0 3797--3820, 2014. doi:10.1109/TIT.2014.2320500

    2014

  18. [26]

    Moore–Penrose inverse

    Wikipedia . Moore–Penrose inverse . https://en.wikipedia.org/wiki/Moore

  19. [27]

    Tonelli’s theorem for non-negative measurable functions

    Wikipedia . Tonelli’s theorem for non-negative measurable functions . https://en.wikipedia.org/wiki/Fubini

  20. [28]

    H., Mangasarian, O

    Wolberg, W. H., Mangasarian, O. L., and Street, W. N. Breast cancer wisconsin (diagnostic) data set. https://archive.ics.uci.edu/ml/datasets/Breast+Cancer+Wisconsin+(Diagnostic), 1993. Accessed: 2025-04-29

    1993

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.