arXiv subjects

Iden Kalemaj

Publications and source records attributed to Iden Kalemaj.

9 recordsLinked to original sources

arXiv2026

Differentially Private Modeling of Disease Transmission within Human Contact Networks

Epidemiologic studies of infectious diseases often rely on models of contact networks to capture the complex interactions that govern disease spread, and ongoing projects aim to vastly increase the scale at which such data can be collected. However, contact networks may include sensitive information, such as sexual relationships or drug use behavior. Protecting individual privacy while maintaining the scientific usefulness of the data is crucial. We propose a privacy-preserving pipeline for disease spread simulation studies based on a sensitive network that integrates differential privacy (DP) with statistical network models such as stochastic block models (SBMs) and exponential random graph models (ERGMs). Our pipeline comprises three steps: (1) compute network summary statistics using \emph{node-level} DP (which corresponds to protecting individuals' contributions); (2) fit a statistical model, like an ERGM, using these summaries, which allows generating synthetic networks reflecting the structure of the original network; and (3) simulate disease spread on the synthetic networks using an agent-based model. We evaluate the effectiveness of our approach using a simple Susceptible-Infected-Susceptible (SIS) disease model under multiple configurations. We compare both numerical results, such as simulated disease incidence and prevalence, as well as qualitative conclusions such as intervention effect size, on networks generated with and without differential privacy constraints. Our experiments are based on egocentric sexual network data from the ARTNet study (a survey about HIV-related behaviors). Our results show that the noise added for privacy is small relative to other sources of error (sampling and model misspecification). This suggests that, in principle, curators of such sensitive data can provide valuable epidemiologic insights while protecting privacy.

cs.CR↗

arXiv2025

Observational Auditing of Label Privacy

Differential privacy (DP) auditing is essential for evaluating privacy guarantees in machine learning systems. Existing auditing methods, however, pose a significant challenge for large-scale systems since they require modifying the training dataset -- for instance, by injecting out-of-distribution canaries or removing samples from training. Such interventions on the training data pipeline are resource-intensive and involve considerable engineering overhead. We introduce a novel observational auditing framework that leverages the inherent randomness of data distributions, enabling privacy evaluation without altering the original dataset. Our approach extends privacy auditing beyond traditional membership inference to protected attributes, with labels as a special case, addressing a key gap in existing techniques. We provide theoretical foundations for our method and perform experiments on Criteo and CIFAR-10 datasets that demonstrate its effectiveness in auditing label privacy guarantees. This work opens new avenues for practical privacy auditing in large-scale production environments.

cs.LG↗

arXiv2025

PrivacyGuard: A Modular Framework for Privacy Auditing in Machine Learning

The increasing deployment of Machine Learning (ML) models in sensitive domains motivates the need for robust, practical privacy assessment tools. PrivacyGuard is a comprehensive tool for empirical differential privacy (DP) analysis, designed to evaluate privacy risks in ML models through state-of-the-art inference attacks and advanced privacy measurement techniques. To this end, PrivacyGuard implements a diverse suite of privacy attack -- including membership inference , extraction, and reconstruction attacks -- enabling both off-the-shelf and highly configurable privacy analyses. Its modular architecture allows for the seamless integration of new attacks, and privacy metrics, supporting rapid adaptation to emerging research advances. We make PrivacyGuard available at https://github.com/facebookresearch/PrivacyGuard.

cs.LG↗

arXiv2024

Differentially Private Conditional Independence Testing

Conditional independence (CI) tests are widely used in statistical data analysis, e.g., they are the building block of many algorithms for causal graph discovery. The goal of a CI test is to accept or reject the null hypothesis that $X \perp \!\!\! \perp Y \mid Z$, where $X \in \mathbb{R}, Y \in \mathbb{R}, Z \in \mathbb{R}^d$. In this work, we investigate conditional independence testing under the constraint of differential privacy. We design two private CI testing procedures: one based on the generalized covariance measure of Shah and Peters (2020) and another based on the conditional randomization test of Candès et al. (2016) (under the model-X assumption). We provide theoretical guarantees on the performance of our tests and validate them empirically. These are the first private CI tests with rigorous theoretical guarantees that work for the general case when $Z$ is continuous.

stat.ML↗

arXiv2023

Counting Distinct Elements in the Turnstile Model with Differential Privacy under Continual Observation

Privacy is a central challenge for systems that learn from sensitive data sets, especially when a system's outputs must be continuously updated to reflect changing data. We consider the achievable error for differentially private continual release of a basic statistic - the number of distinct items - in a stream where items may be both inserted and deleted (the turnstile model). With only insertions, existing algorithms have additive error just polylogarithmic in the length of the stream $T$. We uncover a much richer landscape in the turnstile model, even without considering memory restrictions. We show that every differentially private mechanism that handles insertions and deletions has worst-case additive error at least $T^{1/4}$ even under a relatively weak, event-level privacy definition. Then, we identify a parameter of the input stream, its maximum flippancy, that is low for natural data streams and for which we give tight parameterized error guarantees. Specifically, the maximum flippancy is the largest number of times that the contribution of a single item to the distinct elements count changes over the course of the stream. We present an item-level differentially private mechanism that, for all turnstile streams with maximum flippancy $w$, continually outputs the number of distinct elements with an $O(\sqrt{w} \cdot poly\log T)$ additive error, without requiring prior knowledge of $w$. We prove that this is the best achievable error bound that depends only on $w$, for a large range of values of $w$. When $w$ is small, the error of our mechanism is similar to the polylogarithmic in $T$ error in the insertion-only setting, bypassing the hardness in the turnstile model.

cs.DS↗

arXiv2023

Node-Differentially Private Estimation of the Number of Connected Components

We design the first node-differentially private algorithm for approximating the number of connected components in a graph. Given a database representing an $n$-vertex graph $G$ and a privacy parameter $\varepsilon$, our algorithm runs in polynomial time and, with probability $1-o(1)$, has additive error $\widetilde{O}(\frac{Δ^*\ln\ln n}{\varepsilon}),$ where $Δ^*$ is the smallest possible maximum degree of a spanning forest of $G.$ Node-differentially private algorithms are known only for a small number of database analysis tasks. A major obstacle for designing such an algorithm for the number of connected components is that this graph statistic is not robust to adding one node with arbitrary connections (a change that node-differential privacy is designed to hide): every graph is a neighbor of a connected graph. We overcome this by designing a family of efficiently computable Lipschitz extensions of the number of connected components or, equivalently, the size of a spanning forest. The construction of the extensions, which is at the core of our algorithm, is based on the forest polytope of $G.$ We prove several combinatorial facts about spanning forests, in particular, that a graph with no induced $Δ$-stars has a spanning forest of degree at most $Δ$. With this fact, we show that our Lipschitz extensions for the number of connected components equal the true value of the function for the largest possible monotone families of graphs. More generally, on all monotone sets of graphs, the $\ell_\infty$ error of our Lipschitz extensions is nearly optimal.

cs.DS↗

arXiv2022

Performative Prediction in a Stateful World

Deployed supervised machine learning models make predictions that interact with and influence the world. This phenomenon is called performative prediction by Perdomo et al. (ICML 2020). It is an ongoing challenge to understand the influence of such predictions as well as design tools so as to control that influence. We propose a theoretical framework where the response of a target population to the deployed classifier is modeled as a function of the classifier and the current state (distribution) of the population. We show necessary and sufficient conditions for convergence to an equilibrium of two retraining algorithms, repeated risk minimization and a lazier variant. Furthermore, convergence is near an optimal classifier. We thus generalize results of Perdomo et al., whose performativity framework does not assume any dependence on the state of the target population. A particular phenomenon captured by our model is that of distinct groups that acquire information and resources at different rates to be able to respond to the latest deployed classifier. We study this phenomenon theoretically and empirically.

cs.LG↗

arXiv2021

Sublinear-Time Computation in the Presence of Online Erasures

We initiate the study of sublinear-time algorithms that access their input via an online adversarial erasure oracle. After answering each input query, such an oracle can erase $t$ input values. Our goal is to understand the complexity of basic computational tasks in extremely adversarial situations, where the algorithm's access to data is blocked during the execution of the algorithm in response to its actions. Specifically, we focus on property testing in the model with online erasures. We show that two fundamental properties of functions, linearity and quadraticity, can be tested for constant $t$ with asymptotically the same complexity as in the standard property testing model. For linearity testing, we prove tight bounds in terms of $t$, showing that the query complexity is $\Theta(\log t).$ In contrast to linearity and quadraticity, some other properties, including sortedness and the Lipschitz property of sequences, cannot be tested at all, even for $t=1$. Our investigation leads to a deeper understanding of the structure of violations of linearity and other widely studied properties. We also consider implications of our results for algorithms that are resilient to online adversarial corruptions instead of erasures.

cs.DS↗

arXiv2020

Isoperimetric Inequalities for Real-Valued Functions with Applications to Monotonicity Testing

We generalize the celebrated isoperimetric inequality of Khot, Minzer, and Safra~(SICOMP 2018) for Boolean functions to the case of real-valued functions $f \colon \{0,1\}^d\to\mathbb{R}$. Our main tool in the proof of the generalized inequality is a new Boolean decomposition that represents every real-valued function $f$ over an arbitrary partially ordered domain as a collection of Boolean functions over the same domain, roughly capturing the distance of $f$ to monotonicity and the structure of violations of $f$ to monotonicity. We apply our generalized isoperimetric inequality to improve algorithms for testing monotonicity and approximating the distance to monotonicity for real-valued functions. Our tester for monotonicity has query complexity $\widetilde{O}(\min(r \sqrt{d},d))$, where $r$ is the size of the image of the input function. (The best previously known tester, by Chakrabarty and Seshadhri (STOC 2013), makes $O(d)$ queries.) Our tester is nonadaptive and has 1-sided error. We show a matching lower bound for nonadaptive, 1-sided error testers for monotonicity. We also show that the distance to monotonicity of real-valued functions that are $α$-far from monotone can be approximated nonadaptively within a factor of $O(\sqrt{d\log d})$ with query complexity polynomial in $1/α$ and the dimension $d$. This query complexity is known to be nearly optimal for nonadaptive algorithms even for the special case of Boolean functions. (The best previously known distance approximation algorithm for real-valued functions, by Fattal and Ron (TALG 2010) achieves $O(d\log r)$-approximation.)

cs.DM↗