A new privacy accountant for Gaussian differential privacy measures individual privacy losses.
problem Bounding differential privacy loss for each participant in data analysis.
method Developed a privacy accountant for adaptive compositions of randomised mechanisms using Gaussian differential privacy.
result Provided optimal bounds for the Gaussian mechanism and constructed an approximative individual privacy accountant.
The paper analyzes and proposes methods for privately sharing individual privacy losses using per-instance differential privacy.
problem The standard differential privacy framework provides a worst-case bound that may not accurately reflect individual privacy losses.
method The paper analyzes per-instance differential privacy and proposes methods to privately and accurately publish per-instance privacy losses.
result The methods privately and accurately publish per-instance differential privacy losses with minimal additional privacy cost.
This paper quantifies privacy loss in exploratory data analysis.
problem Privacy loss in exploratory data analysis is often overlooked in privacy budgets.
method Quantitative analysis of privacy loss for statistical functions.
result Privacy loss must be considered in calculating machine learning privacy budgets.
A new method for tighter privacy loss accounting in adaptive analyses.
problem Ensuring individual privacy in adaptive analyses while staying within a privacy budget.
method A personalized privacy loss estimate and a Rényi differential privacy filter.
result Personalized privacy loss accounting can be practical and tighter than existing methods.
BUDS balances privacy and utility by shuffling data, achieving strong privacy with minimal loss.
problem Balancing privacy and utility in crowd-sourced statistical databases.
method One-hot encoding, iterative shuffling, loss estimation, risk minimization.
result Achieves ε=0.02 for privacy, maintaining a privacy bound of ε=ln[t/((n1−1)S)]. Paper improves privacy and utility of SGD with bounded domain and smooth losses.
problem Lack of tight privacy bounds and practical assumptions in DPSGD.
method Rigorous privacy characterization for DPSGD with general L-smooth and non-convex loss functions, tracking privacy loss over iterations.
result Privacy loss converges without convexity assumption for bounded domain, improving utility.
Paper simplifies DP composition for adaptive privacy budgets, enabling better privacy and accuracy in deep learning.
problem Tension between efficiency and flexibility in DP composition theorems.
method Rényi Differential Privacy (RDP) for adaptive privacy budgets, proving simpler composition theorem with smaller constants.
result Practical DP composition for adaptive privacy budgets, enabling better privacy and accuracy in deep learning.
Edgeworth Accountant calculates privacy loss under differential privacy compositions efficiently.
problem Efficiently computing overall privacy loss under composition of private algorithms.
method Analytical approach using f-differential privacy framework and Edgeworth expansion. result Non-asymptotic (ε,δ)-differential privacy bounds with reduced computational cost. Develops a privacy-preserving algorithm for sparse robust regression.
problem Privacy-preserving machine learning for sparse robust regression.
method Develops FRAPPE algorithm for non-smooth loss under differential privacy.
result Achieves better privacy and statistical accuracy trade-off.
Study on privacy leakage in noisy gradient descent algorithms.
problem Information leakage of iterative randomized learning algorithms about training data.
method Analyzes the dynamics of Rényi differential privacy loss in noisy gradient descent algorithms.
result Privacy loss converges exponentially fast for smooth and strongly convex loss functions.
The paper improves privacy accounting for discrete-valued mechanisms and the subsampled Gaussian mechanism.
problem Improving the accuracy and efficiency of differential privacy accounting for discrete outputs.
method Uses fast Fourier transform (FFT) for rigorous error analysis and accounting of privacy loss.
result Provides strict lower and upper bounds for (ε,δ)-values, demonstrating up to 75% reduction in noise variance. New analysis shows SGD with noise doesn't leak more privacy with more iterations.
problem Privacy loss in noisy SGD with more iterations.
method Privacy Amplification by Iteration and Sampled Gaussian Mechanism.
result Privacy loss remains constant after a burn-in period, not increasing with more iterations.
Differential privacy is a strong notion for privacy that can be used to prove formal guarantees, in terms of a privacy budget, ε, about how much information is leaked by a mechanism. However, implementations of privacy-preserving machine learning often select large values of ε in order to get acceptable utility of …
Log-Loss scores expose membership privacy breaches.
problem Privacy leakage from statistical aggregates like Log-Loss scores.
method Proved that Log-Loss scores enable full accuracy membership inference in a single query.
result Complete membership privacy breach is possible with Log-Loss scores.
Paper improves privacy-preserving measurement of advertising incrementality.
problem Privacy degradation in randomized lift tests for advertising measurement.
method Formulates a robust causal decision problem under signal losses, projecting clean worlds onto incrementality.
result Sharp decision frontier shows valid certification or rejection outside the frontier.
DPlis improves privacy in deep learning models by smoothing loss functions.
problem Privacy leakage in deep learning models trained on private data and low model performance.
method DPlis constructs a smooth loss function to favor noise-resilient models.
result DPlis effectively boosts model quality and training stability under privacy constraints.
Develops a computationally tractable high-dimensional differential privacy estimator.
problem Differential privacy in high dimensions is computationally intractable.
method Combines high-dimensional robust statistics with differential privacy techniques.
result A computationally tractable algorithm with dimension-independent privacy loss.
Constructs models to meet any DP requirement without retraining.
problem Evolving privacy requirements at inference time.
method Two post-processing techniques: random selection and linear combination.
result Final private models satisfying any target privacy parameter.
Paper introduces privacy-preserving inventory policy learning for feature-based newsvendor with unknown demand.
problem Privacy-preserving inventory policy learning for feature-based newsvendor with unknown demand distribution and nonsmooth loss function.
method Developed a clipped noisy gradient descent algorithm based on convolution smoothing for optimal inventory estimation within f-differential privacy framework.
result Achieved privacy-preserving optimal inventory policy with provable privacy guarantees and desirable statistical precision.
This paper sharpens privacy guarantees for high-dimensional PCA under differential privacy.
problem Understanding the exact privacy loss in high-dimensional PCA with differential privacy.
method Analyzes the exponential mechanism in a model-free setting for high-dimensional PCA.
result Sharp utility and privacy characterizations in high dimensions show the difficulty of detecting a target individual's presence.
Improved privacy analysis for stochastic gradient descent.
problem Analyzing privacy leakage in noisy stochastic gradient descent.
method Modeling Rényi divergence dynamics with Langevin diffusions, proving exponential privacy loss convergence for smooth and strongly convex objectives.
result Privacy loss converges exponentially fast for smooth and strongly convex objectives under constant step size.
New method reduces privacy impact on model accuracy for underrepresented groups.
problem Privacy mechanisms disproportionately affect underrepresented groups in machine learning models.
method Proposes DPSGD-F, a modified DPSGD that adjusts group contributions based on clipping bias.
result DPSGD-F removes disparate impact of differential privacy on model accuracy for protected groups.
SMOTE-DP enhances synthetic data privacy without sacrificing utility.
problem Balancing privacy and utility in synthetic data generation.
method Integrating SMOTE with differential privacy mechanisms.
result SMOTE-DP produces synthetic data that is both private and useful.
The Gradient Boosting Decision Tree (GBDT) is a popular machine learning model for various tasks in recent years. In this paper, we study how to improve model accuracy of GBDT while preserving the strong guarantee of differential privacy. Sensitivity and privacy budget are two key design aspects for the effectiveness o…
Improved privacy-preserving methods for convex optimization with heavy-tailed data.
problem Privacy-preserving optimization of convex functions with heavy-tailed data.
method Developed algorithms for private mean estimation and convex optimization under concentrated differential privacy constraints.
result Achieved improved upper bounds on excess population risk for convex and strongly convex loss functions.
We develop a novel approximate Bayesian computation (ABC) framework, ABCDP, that produces differentially private (DP) and approximate posterior samples. Our framework takes advantage of the Sparse Vector Technique (SVT), widely studied in the differential privacy literature. SVT incurs the privacy cost only when a cond…
New filters match advanced composition for adaptive privacy, with practical constants.
problem Limitations of existing adaptive composition methods.
method Constructed new filters and odometers that match advanced composition rates, including constants.
result Achieved fully adaptive privacy with practical filters and odometers.
Paper improves differential privacy analysis for machine learning.
problem Quantifying privacy leakage in noisy gradient descent.
method Shifted interpolation in f-differential privacy. result First exact privacy analysis for strongly convex optimization.
Paper extends FFT-based differential privacy method to heterogeneous compositions.
problem Computing accurate differential privacy guarantees for mixed mechanisms.
method Uses Fast Fourier Transform (FFT) for error analysis and parameter selection.
result Provides tighter bounds for heterogeneous compositions compared to homogeneous cases.
Enhanced privacy, utility, and efficiency through MUST subsampling.
problem Balancing privacy, utility, and computational efficiency in data analysis.
method MUltistage Sampling Technique (MUST) for privacy amplification in differential privacy.
result MUST offers stronger privacy guarantees (ϵ) than one-stage subsampling methods while maintaining similar utility and computational efficiency. The paper introduces a privacy-preserving method for estimating treatment effects that maintains accuracy.
problem Estimating heterogeneous treatment effects in sensitive data while protecting privacy.
method A general meta-algorithm for CATE estimation with differential privacy guarantees, using sample splitting and parallel composition.
result The meta-algorithm maintains accuracy even with differential privacy, showing that most accuracy loss is due to variance increase.
RDP-GAN improves GAN privacy by adding random noises to loss function.
problem Protecting sensitive information in GANs while maintaining quality of generated samples.
method Integrates Rényi-differential privacy into GAN training process by adding random noises to loss function.
result Achieves better privacy protection with high-quality samples compared to existing methods.
Paper relaxes SGD privacy and generalization guarantees for non-smooth convex losses.
problem Privacy and generalization in SGD for non-smooth convex losses.
method Relaxes Lipschitz and strong smoothness assumptions to Hölder smoothness, proving (ε,δ)-DP and optimal excess risk. result Noisy SGD with α-Hölder smooth losses achieves optimal excess risk with linear gradient complexity for α≥1/2. We consider the problem of reinforcing federated learning with formal privacy guarantees. We propose to employ Bayesian differential privacy, a relaxation of differential privacy for similarly distributed data, to provide sharper privacy loss bounds. We adapt the Bayesian privacy accounting method to the federated sett…
New method assesses individual training points' privacy risk without retraining.
problem Privacy vulnerability of individual training points in membership inference attacks.
method Derives a closed-form decomposition of individual black-box MIA vulnerability, extending to deep networks.
result Proposes a surrogate score operating on last-layer representations that requires only a single trained model.
Paper proves privacy guarantees for shuffled and online PNSGD, reducing noise over time.
problem Privacy amplification in shuffled and online PNSGD settings.
method Iterative analysis of PNSGD with hidden updates, proving privacy guarantees for shuffled and online settings.
result Privacy guarantees for shuffled and online PNSGD with reduced noise over time.
Differential privacy allows quantifying privacy loss resulting from accessing sensitive personal data. Repeated accesses to underlying data incur increasing loss. Releasing data as privacy-preserving synthetic data would avoid this limitation, but would leave open the problem of designing what kind of synthetic data. W…
New framework for DP-SMO with near-optimal privacy-loss trade-off.
problem Optimal trade-off between privacy and population loss in DP-SMO.
method General framework using Phased-ERM method and black-box optimization.
result Near-linear time algorithms with near-optimal guarantees.
New mechanisms improve differential privacy for scalar queries.
problem Improving differential privacy for scalar, real-valued query functions.
method Mixing multiple Gaussian distributions to satisfy differential privacy.
result Mechanisms yield lower noise amplitudes and variances compared to the analytic Gaussian mechanism.
Link prediction (LP) algorithms propose to each node a ranked list of nodes that are currently non-neighbors, as the most likely candidates for future linkage. Owing to increasing concerns about privacy, users (nodes) may prefer to keep some of their connections protected or private. Motivated by this observation, our …
Normalization layers improve the accuracy of Differentially Private training of deep neural networks.
problem Reduced accuracy in deep neural networks with Differentially Private training.
method Proposed a novel method for integrating batch normalization with Differentially Private Stochastic Gradient Descent (DPSGD) without additional privacy loss.
result Training deeper networks with better utility-privacy trade-off is possible.
Iteratively reweighted least squares (IRLS) is a widely-used method in machine learning to estimate the parameters in the generalised linear models. In particular, IRLS for L1 minimisation under the linear model provides a closed-form solution in each step, which is a simple multiplication between the inverse of the we…
Second-order methods improve differential privacy in convex optimization.
problem Improving differential privacy in convex optimization.
method Developed a private variant of the regularized cubic Newton method for strongly convex loss functions.
result Achieves quadratic convergence and optimal excess loss for strongly convex loss functions.
This paper bounds min-entropy leakage for Blowfish privacy using graph symmetries.
problem Bounding min-entropy leakage for Blowfish privacy mechanisms.
method Organizing analysis over symmetrical partitions corresponding to orbits of graph automorphism groups.
result Demonstrates a construction meeting the bound with asymptotic equality, showing tightness.
This work optimizes mean estimation under varying user privacy demands.
problem Mean estimation with heterogeneous privacy levels.
method Proposes an algorithm that is minimax optimal and has near-linear run-time.
result Privacy requirements of the most stringent users dictate overall error rates.
This work optimizes mean estimation under varying privacy constraints.
problem Mean estimation with heterogeneous privacy constraints.
method Proposes an algorithm for mean estimation under different privacy levels for users.
result Shows a saturation phenomenon in performance as privacy levels are relaxed.
We present a data-driven framework called generative adversarial privacy (GAP). Inspired by recent advancements in generative adversarial networks (GANs), GAP allows the data holder to learn the privatization mechanism directly from the data. Under GAP, finding the optimal privacy mechanism is formulated as a constrain…
New privacy bounds for DP-SGD's last iterate, even with cyclic sampling.
problem Privacy of the last iterate in DP-SGD with cyclic sampling.
method Established new RDP upper bounds for the last iterate under realistic assumptions.
result Privacy bounds for DP-SGD's last iterate with cyclic sampling and clipping, even for nonconvex losses.