arXiv subjects

Martin Dunsche

Publications and source records attributed to Martin Dunsche.

6 recordsLinked to original sources

arXiv2026

Sequential Auditing for f-Differential Privacy

We present new auditors to assess Differential Privacy (DP) of an algorithm based on output samples. Such empirical auditors are common to check for algorithmic correctness and implementation bugs. Most existing auditors are batch-based or targeted toward the traditional notion of $(\varepsilon,\delta)$-DP; typically both. In this work, we shift the focus to the highly expressive privacy concept of $f$-DP, in which the entire privacy behavior is captured by a single tradeoff curve. Our auditors detect violations across the full privacy spectrum with statistical significance guarantees, which are supported by theory and simulations. Most importantly, and in contrast to prior work, our auditors do not require a user-specified sample size as an input. Rather, they adaptively determine a near-optimal number of samples needed to reach a decision, thereby avoiding the excessively large sample sizes common in many auditing studies. This reduction in sampling cost becomes especially beneficial for expensive training procedures such as DP-SGD. Our method supports both whitebox and blackbox settings and can also be executed in one-run frameworks.

cs.CR↗

arXiv2025

Differentially private testing for relevant dependencies in high dimensions

We investigate the problem of detecting dependencies between the components of a high-dimensional vector. Our approach advances the existing literature in two important respects. First, we consider the problem under privacy constraints. Second, instead of testing whether the coordinates are pairwise independent, we are interested in determining whether certain pairwise associations between the components (such as all pairwise Kendall's $\tau$ coefficients) do not exceed a given threshold in absolute value. Considering hypotheses of this form is motivated by the observation that in the high-dimensional regime, it is rare and perhaps impossible to have a null hypothesis that can be modeled exactly by assuming that all pairwise associations are precisely equal to zero. The formulation of the null hypothesis as a composite hypothesis makes the problem of constructing tests already non-standard in the non-private setting. Additionally, under privacy constraints, state of the art procedures rely on permutation approaches that are rendered invalid under a composite null. We propose a novel bootstrap based methodology that is especially powerful in sparse settings, develop theoretical guarantees under mild assumptions and show that the proposed method enjoys good finite sample properties even in the high privacy regime. Additionally, we present applications in medical data that showcase the applicability of our methodology.

math.ST↗

arXiv2025

SILENT: A New Lens on Statistics in Software Timing Side Channels

Cryptographic research takes software timing side channels seriously. Approaches to mitigate them include constant-time coding and techniques to enforce such practices. However, recent attacks like Meltdown [42], Spectre [37], and Hertzbleed [70] have challenged our understanding of what it means for code to execute in constant time on modern CPUs. To ensure that assumptions on the underlying hardware are correct and to create a complete feedback loop, developers should also perform \emph{timing measurements} as a final validation step to ensure the absence of exploitable side channels. Unfortunately, as highlighted by a recent study by Jancar et al. [30], developers often avoid measurements due to the perceived unreliability of the statistical analysis and its guarantees. In this work, we combat the view that statistical techniques only provide weak guarantees by introducing a new algorithm for the analysis of timing measurements with strong, formal statistical guarantees, giving developers a reliable analysis tool. Specifically, our algorithm (1) is non-parametric, making minimal assumptions about the underlying distribution and thus overcoming limitations of classical tests like the t-test, (2) handles unknown data dependencies in measurements, (3) can estimate in advance how many samples are needed to detect a leak of a given size, and (4) allows the definition of a negligible leak threshold $\Delta$, ensuring that acceptable non-exploitable leaks do not trigger false positives, without compromising statistical soundness. We demonstrate the necessity, effectiveness, and benefits of our approach on both synthetic benchmarks and real-world applications.

cs.CR↗

arXiv2025

General-Purpose $f$-DP Estimation and Auditing in a Black-Box Setting

In this paper we propose new methods to statistically assess $f$-Differential Privacy ($f$-DP), a recent refinement of differential privacy (DP) that remedies certain weaknesses of standard DP (including tightness under algorithmic composition). A challenge when deploying differentially private mechanisms is that DP is hard to validate, especially in the black-box setting. This has led to numerous empirical methods for auditing standard DP, while $f$-DP remains less explored. We introduce new black-box methods for $f$-DP that, unlike existing approaches for this privacy notion, do not require prior knowledge of the investigated algorithm. Our procedure yields a complete estimate of the $f$-DP trade-off curve, with theoretical guarantees of convergence. Additionally, we propose an efficient auditing method that empirically detects $f$-DP violations with statistical certainty, merging techniques from non-parametric estimation and optimal classification theory. Through experiments on a range of DP mechanisms, we demonstrate the effectiveness of our estimation and auditing procedures.

cs.CR↗

arXiv2022

Lower Bounds for R\'enyi Differential Privacy in a Black-Box Setting

We present new methods for assessing the privacy guarantees of an algorithm with regard to R\'enyi Differential Privacy. To the best of our knowledge, this work is the first to address this problem in a black-box scenario, where only algorithmic outputs are available. To quantify privacy leakage, we devise a new estimator for the R\'enyi divergence of a pair of output distributions. This estimator is transformed into a statistical lower bound that is proven to hold for large samples with high probability. Our method is applicable for a broad class of algorithms, including many well-known examples from the privacy literature. We demonstrate the effectiveness of our approach by experiments encompassing algorithms and privacy enhancing methods that have not been considered in related works.

cs.CR↗

arXiv2021

Multivariate Mean Comparison under Differential Privacy

The comparison of multivariate population means is a central task of statistical inference. While statistical theory provides a variety of analysis tools, they usually do not protect individuals' privacy. This knowledge can create incentives for participants in a study to conceal their true data (especially for outliers), which might result in a distorted analysis. In this paper we address this problem by developing a hypothesis test for multivariate mean comparisons that guarantees differential privacy to users. The test statistic is based on the popular Hotelling's $t^2$-statistic, which has a natural interpretation in terms of the Mahalanobis distance. In order to control the type-1-error, we present a bootstrap algorithm under differential privacy that provably yields a reliable test decision. In an empirical study we demonstrate the applicability of this approach.

stat.ME↗