Infinite Quantiles Inference under Local Differential Privacy
Authors: Yi Liu, Qirui Hu
Organizations: Department of Mathematics and Statistics, York University · School of Statistics and Data Science, Shanghai University of Finance and Economics. · Department of Mathematics, Ruhr University Bochum
We study simultaneous quantile inference under local differential privacy for a number of target probabilities that grows with the sample size. Each user participates once, so learning unknown quantile locations and calibrating their joint uncertainty must share the same private information. We fuse complementary cumulative distribution function (CDF) scores from a single tree report using their exact covariance; the variance at the correct anchor is no larger than that of either score, and anchor error incurs only quadratic excess variance. Three consecutive cohorts then localize the quantiles, estimate their densities, and supply fresh scores for a one-step correction and joint multiplier calibration. Under mild conditions, a uniform representation controls localization and density errors and yields asymptotically valid simultaneous coverage. A fixed-privacy regime permits every polynomially growing target grid. Simulations show consistent same-report precision gains, while comparisons with CDF-band inversion expose the sample cost of learning target locations. Census microdata illustrates how the resulting joint intervals support wage-percentile ratios.
Figures & tables
k
n
Fused
Balanced
kRR
31
160000
0.932; 0.0897
0.926; 0.1147
0.918; 0.1545
31
320000
0.940; 0.0602
0.944; 0.0766
0.928; 0.0949
63
320000
0.944; 0.0698
0.938; 0.0903
0.892; 0.2052
63
640000
0.948; 0.0490
0.936; 0.0625
0.902; 0.1340
127
640000
0.948; 0.0561
0.952; 0.0716
0.844; 0.3538
127
1280000
0.950; 0.0386
0.952; 0.0494
0.860; 0.2038
Table 1: End-to-end sample-size scaling at ϵ=1 under N(0,1) . Fused is the complete procedure in Algorithm 1 ; Balanced changes its tree decoder, and kRR changes its reporting channel. All methods use the same three-cohort allocation and public bracket. Each entry is simultaneous coverage followed by pooled RMSE; lower RMSE is better.
ϵ
Fused
Balanced
kRR
0.5
0.944; 0.1490
0.932; 0.2193
0.874; 1.1007
1.0
0.944; 0.0698
0.938; 0.0903
0.892; 0.2052
2.0
0.956; 0.0326
0.932; 0.0402
0.934; 0.0474
Table 2: End-to-end privacy comparison at n=320000 and k=63 under N(0,1) . Each method entry is simultaneous coverage followed by pooled RMSE.
Distribution
Fused
Balanced
kRR
Normal
0.948; 0.0490
0.936; 0.0625
0.902; 0.1340
t3
0.948; 0.0483
0.932; 0.0597
0.862; 0.1182
Contamination
0.938; 0.0842
0.932; 0.1053
0.844; 0.2377
Table 3: End-to-end distribution comparison at n=640000 , k=63 , and ϵ=1 . Each method entry is simultaneous coverage followed by pooled RMSE.
n
k
Fused three-cohort
CDF inversion ( d=k )
40000
31
0.942; 0.2704; 1.1112
1.000; 0.1064; 0.8216
160000
31
0.950; 0.1134; 0.5146
0.994; 0.0512; 0.5067
320000
63
0.918; 0.0809; 0.4326
0.992; 0.0395; 0.3556
Table 4: Complete procedures under N(0,1) on [−3,3] , with one ε=1 report per user. Each entry is (simultaneous coverage; RMSE; average full width). Both methods use all n users; the fused three-cohort method learns quantile locations and densities. Lower RMSE and width are preferable at comparable coverage.
n
k
All runs
No clipping
Clipping
40000
31
0.942[0.918,0.959]
349/370
122/130
160000
31
0.950[0.927,0.966]
452/476
23/24
320000
63
0.918[0.891,0.939]
450/490
9/10
Table 5: Fused three-cohort inference on [−3,3] under N(0,1) at ε=1 . All-run entries give coverage and its 95% Wilson interval across 500 repetitions. The last two columns give covered repetitions divided by repetitions in that density-clipping group.
n
k
Method
Coverage
RMSE
Width
160000
31
Fused three-cohort
0.932
0.0897
0.4777
Student–Wishart
0.938
0.0764
0.6367
Student–Bonferroni
0.948
0.0764
0.6432
320000
63
Fused three-cohort
0.944
0.0698
0.4178
Student–Wishart
0.934
0.0662
0.6165
Student–Bonferroni
0.946
0.0662
0.6240
Table 6: Corrected block procedures versus fused three-cohort inference ( Algorithm 1 ) at ϵ=1 under N(0,1) . Restarted methods use J=8 and balanced tree-OUE; the complete designs differ as specified in the text. Width is the mean full interval width.
τ
Empirical benchmark
Private estimate
95% simultaneous interval
0.10
14.40
13.44
[10.85, 16.66]
0.25
20.50
18.96
[16.27, 22.10]
0.50
34.17
30.71
[26.72, 35.30]
0.75
56.52
51.50
[43.69, 60.71]
0.90
90.30
100.37
[76.54, 131.61]
Table 7: Hourly-wage quantiles in 2024 dollars. Intervals use nominal 95% simultaneous calibration under the continuous working model.
Figure 1: Private hourly-wage quantiles and nominal 95% simultaneous intervals for the 19-point ACS grid in Section 6 . The left panel uses a logarithmic wage axis; the right enlarges the middle quantiles on a linear axis. Black triangles are empirical file benchmarks. At the 40th percentile, the benchmark lies above its interval. Calibration targets the continuous working-model quantiles; the plotted benchmarks describe the fixed released file.
With the development of big data and machine learning, privacy concerns have become increasingly critical, especially when handling heterogeneous datasets containing sensitive personal information. Differential privacy provides a rigorous framework for safeguarding individual privacy while enabling meaningful statistical analysis. In this paper, we propose a differentially private quantile regression method for high-dimensional data in a distributed setting. Quantile regression is a powerful and robust tool for modeling the relationships between the covariates and responses in the presence of outliers or heavy-tailed distributions. To address the computational challenges due to the non-smoothness of the quantile loss function, we introduce a Newton-type transformation that reformulates the quantile regression task into an ordinary least squares problem. Building on this, we develop a differentially private estimation algorithm with iterative updates, ensuring both near-optimal statistical accuracy and formal privacy guarantees. For inference, we further propose a differentially private debiased estimator, which enables valid confidence interval construction and hypothesis testing. Additionally, we propose a communication-efficient and differentially private bootstrap for simultaneous hypothesis testing in high-dimensional quantile regression, suitable for distributed settings with both small and abundant local data. Extensive simulations demonstrate the robustness and effectiveness of our methods in practical scenarios.
Ziliang Shen, Caixing Wang, Shaoli Wang +1
School of Statistics and Data Science, Shanghai University of Finance and Economics · School of Statistics and Data Science, Southeast University · School of Statistics and Data Science, Shanghai University of International Business and Economics
Many differentially private (DP) data release systems either output DP synthetic data and leave analysts to perform inference as usual, which can lead to severe miscalibration, or output a DP point estimate without a principled way to do uncertainty quantification. This paper develops a clean and tractable middle ground for exponential families: release only DP sufficient statistics, then perform noise-calibrated likelihood-based inference and optional parametric synthetic data generation as post-processing. Our contributions are: (1) a general recipe for approximate-DP release of clipped sufficient statistics under the Gaussian mechanism; (2) asymptotic normality, explicit variance inflation, and valid Wald-style confidence intervals for the plug-in DP MLE; (3) a noise-aware likelihood correction that is first-order equivalent to the plug-in but supports bootstrap-based intervals; and (4) a matching minimax lower bound showing the privacy distortion rate is unavoidable. The resulting theory yields concrete design rules and a practical pipeline for releasing DP synthetic data with principled uncertainty quantification, validated on three exponential families and real census data.
Amir Asiaee, Samhita Pal
Department of Biostatistics, Vanderbilt University Medical Center, Nashville, TN, USA
We develop inference for manifold-valued population minimizers when each observation belongs to a different participant and only locally private messages reach the analyst. The method releases randomized tangent gradients and combines them through Riemannian stochastic approximation and Polyak-Ruppert averaging. Directly inserting a private data surrogate into a nonlinear loss can shift its population target, whereas conditional centring of the released gradient preserves the first-order equation. We introduce symmetric-pair regression (SPR) to estimate the asymptotic variance from the same private messages used for point estimation, without holding out participants or requesting a second release. We prove the central limit theorem and consistency of the fully transcript-based sandwich covariance and intrinsic Wald region under local differential privacy. Simulations across various statistical problems and manifolds support the predicted decrease in estimation error and near-nominal coverage under moderate privacy. An application to NHANES anthropometric data illustrates private estimation of a leading body-size direction and its uncertainty.
Xiaotian Chang, Yangdi Jiang, Qirui Hu
School of Physical and Mathematical Sciences, Nanyang Technological University · School of Statistics and Data Science, Shanghai University of Finance and Economics.