Paper explores how to design federated learning protocols that benefit all participants while maintaining privacy.
problem Privacy concerns undermine the accuracy benefits of federated learning in privacy-sensitive domains.
method The paper provides conditions for mutually beneficial federated learning protocols and designs protocols that maximize total utility and accuracy.
result The paper demonstrates that federated learning can be designed to be mutually beneficial, striking a balance between privacy and model accuracy.
We propose a cloud-based filter trained to block third parties from uploading privacy-sensitive images of others to online social media. The proposed filter uses Distributed One-Class Learning, which decomposes the cloud-based filter into multiple one-class classifiers. Each one-class classifier captures the properties…
Paper tackles noisy S-D data for classification.
problem Learning from noisy Similar (S) and Dissimilar (D) pairs.
method Proposes two algorithms to learn from noisy S-D data under two noise models.
result Noise-informed algorithms outperform noise-blind baselines.
Fine-tuning LLMs on privacy-sensitive data introduces privacy risk, and synthetic data audits can quantify this risk.
problem Fine-tuning LLMs on privacy-sensitive data introduces privacy risk.
method Generate synthetic canaries via high-temperature sampling from LLMs.
result Synthetic canaries are high-influence outliers that ensure strong audits.
We study the problem of data release with privacy, where data is made available with privacy guarantees while keeping the usability of the data as high as possible --- this is important in health-care and other domains with sensitive data. In particular, we propose a method of masking the private data with privacy guar…
AdapTable adapts tabular models to shifts without source data, improving HELOC performance.
problem Distribution shifts in tabular data threaten model performance.
method Shift-aware uncertainty calibrator and label distribution handler.
result Up to 16% improvement on HELOC dataset.
DPDMs use DP-SGD to generate private synthetic data.
problem Generating private synthetic data from sensitive datasets.
method Introduced DP-SGD for DMs, investigated DM parameterization and sampling, proposed noise multiplicity.
result Achieved state-of-the-art performance in image generation benchmarks.
Random projection (RP) is a classical technique for reducing storage and computational costs. We analyze RP-based approximations of convex programs, in which the original optimization problem is approximated by the solution of a lower-dimensional problem. Such dimensionality reduction is essential in computation-limite…
Machine learning has been widely applied to various applications, some of which involve training with privacy-sensitive data. A modest number of data breaches have been studied, including credit card information in natural language data and identities from face dataset. However, most of these studies focus on supervise…
Latent Noise Injection improves synthetic data generation for privacy and statistical alignment.
problem Slow convergence of generative models in high-dimensional settings.
method Latent Noise Injection using Masked Autoregressive Flows (MAF).
result Synthetic data closely reflects the underlying distribution, especially in high-dimensional settings.
We consider the problem of fitting a linear model to data held by individuals who are concerned about their privacy. Incentivizing most players to truthfully report their data to the analyst constrains our design to mechanisms that provide a privacy guarantee to the participants; we use differential privacy to model in…
Early stopping improves generalization in overparameterized diffusion models.
problem Understanding and optimizing generalization in overparameterized diffusion models.
method Revisiting diffusion models, showing generalization occurs before memorization, and developing a phase diagram.
result Generalization time scales with dataset size, supporting early-stopping criteria.
Study evaluates federated learning with differential privacy on MIMIC-III, improving model performance with careful parameter tuning.
problem Training machine learning models on privacy-sensitive data sets locked in healthcare facilities.
method Extensive evaluation of federated and differential privacy techniques on MIMIC-III dataset, analyzing various parameters.
result Careful parameter tuning is crucial for federated learning with differential privacy, especially for data distribution and communication strategies.
In this paper, we address the problem of data reconstruction from privacy-protected templates, based on recent concept of sparse ternary coding with ambiguization (STCA). The STCA is a generalization of randomization techniques which includes random projections, lossy quantization, and addition of ambiguization noise t…
The paper provides Gaussian approximations for decentralized Federated Learning.
problem Lack of asymptotic statistical guarantees for local SGD in Federated Learning.
method Two generalized Gaussian approximation results for local SGD trajectories.
result Valid multiplier bootstrap procedures and Gaussian bootstrap-based tests for detecting adversarial attacks.
Synthetic tabular data improves privacy while maintaining model performance.
problem Protecting privacy in synthetic data generation for machine learning.
method Deep generative models for tabular data, emphasizing privacy and model performance.
result Deep generative models enhance synthetic data generation for tabular datasets.
In this work we address the practical challenges of training machine learning models on privacy-sensitive datasets by introducing a modular approach that minimizes changes to training algorithms, provides a variety of configuration strategies for the privacy mechanism, and then isolates and simplifies the critical logi…
The workshop focuses on AI principles for structured data.
problem Using AI on structured data for decision-making.
method Addressing principles of privacy, accountability, interpretability, robustness, and reasoning.
result Designing approaches to use structured data for reliable decisions.
We study the problem of privacy-preserving machine learning (PPML) for ensemble methods, focusing our effort on random forests. In collaborative analysis, PPML attempts to solve the conflict between the need for data sharing and privacy. This is especially important in privacy sensitive applications such as learning pr…
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.
Novel algorithm for private analysis of fMRI data.
problem Privacy concerns in collaborative neuroimaging data analysis.
method Differentially private decentralized ICA algorithm using correlated noise.
result Algorithm outperforms existing methods on synthetic and real neuroimaging datasets.
New synthetic data analysis reveals high type 1 error rates.
problem Analyzing synthetic data for inference raises significant methodological challenges.
method Developed statistical inference tools and conducted a simulation study.
result Type 1 error rates are unacceptably high in synthetic data analysis.
SAPAG attacks distributed learning by reconstructing true training data from gradients.
problem Privacy attacks on distributed learning systems through gradients.
method SAPAG uses a Gaussian kernel-based gradient difference distance measure.
result SAPAG can reconstruct training data on various DNNs and at different training phases.
Electronic medical records (EMRs) supports the development of machine learning algorithms for predicting disease incidence, patient response to treatment, and other healthcare events. But insofar most algorithms have been centralized, taking little account of the decentralized, non-identically independently distributed…
Text-based analysis methods allow to reveal privacy relevant author attributes such as gender, age and identify of the text's author. Such methods can compromise the privacy of an anonymous author even when the author tries to remove privacy sensitive content. In this paper, we propose an automatic method, called Adver…
Differential privacy mechanism design has traditionally been tailored for a scalar-valued query function. Although many mechanisms such as the Laplace and Gaussian mechanisms can be extended to a matrix-valued query function by adding i.i.d. noise to each element of the matrix, this method is often suboptimal as it for…
Generative models help debug data issues in private, decentralized datasets.
problem Manual data inspection is impractical for private, decentralized datasets.
method Differentially private federated generative models.
result Generative models can effectively debug data issues in private datasets.
New algorithm for federated learning without a central server.
problem Federated learning in unidirectional trust social networks.
method Central Server Free Federated Learning (OPS) method.
result Users can benefit from communication with trusted users in federated learning.
A new method for unlearning trained models without needing the original data.
problem Lack of access to original training data for privacy-preserving unlearning.
method Uses a surrogate dataset to approximate statistical properties and calibrates noise based on statistical distance.
result Effective unlearning of trained models with strong privacy guarantees, even without access to the original data.
FedUA trains UA models privately without raw data.
problem Training UA models requires raw user data, compromising privacy.
method Federated learning framework for privacy-preserving UA model training.
result FedUA reliably rejects unseen user data at high true positive rates.
New methods protect privacy while providing accurate prediction sets.
problem Privacy-preserving conformal prediction for untrusted aggregators.
method Two LDP approaches: k-ary randomized response and binary search response.
result Finite-sample coverage guarantees and robust coverage under randomization.
Generative text classifiers are most vulnerable to membership inference attacks.
problem Privacy threat from Membership Inference Attacks on generative text classifiers.
method Comprehensive empirical evaluation of generative, discriminative, and pseudo-generative classifiers across various datasets.
result Generative classifiers explicitly modeling P(X,Y) are most vulnerable to membership leakage. Paper proposes a new method to assess synthetic data generators.
problem Assessing the quality of implicit generative models.
method Kernelised Stein Statistic (KSD) test based on non-parametric Stein operator.
result Improved power performance compared to existing approaches.
Paper improves speaker verification with federated learning and differential privacy.
problem Improving speaker verification accuracy using private data.
method Combining federated learning and differential privacy to train an auxiliary model that predicts vocal characteristics.
result 6% relative improvement in equal error rate over a baseline system.
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…
Improves GBDT accuracy with differential privacy.
problem Balancing privacy and accuracy in GBDT models.
method Adaptive gradient control and novel boosting framework for privacy budget allocation.
result Achieves better model accuracy with differential privacy.
This paper tackles energy-efficient machine learning on low-power devices.
problem Energy consumption in machine learning due to data communication.
method Dynamic averaging for integer exponential families on low-power processors.
result Achieves comparable model quality with significantly less communication and energy.
Paper develops methods for fair insurance pricing without direct access to sensitive attributes.
problem Fairness in insurance pricing with restricted access to sensitive attributes.
method Develops statistical methods for estimating discrimination-free premiums using privatized sensitive attributes.
result The proposed methods enable fair insurance pricing while respecting privacy and regulatory constraints.
SplitEasy trains ML models on mobile devices without server data transfer.
problem Training complex DL models on resource-limited mobile devices.
method Split learning approach where sensitive layers are trained locally, computationally intensive layers on server.
result SplitEasy trains models on mobile devices with minimal data transfer, near-constant time per sample.
Sparse attention model reduces long-context inference time with exponential accuracy guarantees.
problem Efficiently processing long-context queries in large language models.
method Formalizes attention as a projection onto key vectors, analyzes entropic relaxation, and introduces Vashista Sparse Attention.
result Sparse attention concentrates on a constant-size active face, leading to exponential decay of inactive tokens' mass and linear scaling of active face error.
Privacy-preserving machine learning methods add randomness, leading to varying predictions.
problem Privacy-preserving machine learning methods add randomness, leading to varying predictions.
method The study analyzes three DP-ensuring algorithms: output perturbation, objective perturbation, and DP-SGD.
result The degree of predictive multiplicity rises as the level of privacy increases, and is unevenly distributed across individuals and demographic groups.
New method can infer training data from deep neural networks with high success rates.
problem Model inversion attacks on deep neural networks pose privacy risks.
method Generative model-inversion attack using GANs and partial public information.
result Significant improvement in identifying private training data from deep models.
Access to sufficient annotated data is a common challenge in training deep neural networks on medical images. As annotating data is expensive and time-consuming, it is difficult for an individual medical center to reach large enough sample sizes to build their own, personalized models. As an alternative, data from all …
Bayesian mixture models inference in federated learning
problem Inference of Bayesian mixture models in federated learning
method Consensus Monte Carlo approach
result Recovery of small clusters with greater accuracy than standard MCMC
DP-CDA generates synthetic data to enhance privacy in high-dimensional datasets.
problem Privacy concerns in anonymized datasets, especially in high-dimensional data.
method Randomized mixing of privacy-sensitive data in a class-specific manner with carefully tuned randomness.
result DP-CDA provides stronger privacy guarantees compared to existing methods, maintaining utility.
A two-stage optimization framework reduces label noise in federated learning.
problem Label noise from noisy clients degrades federated learning model performance.
method MaskedOptim framework: detects noisy clients, corrects labels, and aggregates models robustly.
result Our framework improves model robustness and data quality in federated learning.
Method generates intermediate domains to align source and target domains.
problem Challenges of domain adaptation with significant domain divergence.
method Progressive domain augmentation via domain interpolation and multiple subspace alignment.
result Achieves state-of-the-art performance on multiple domain adaptation tasks.
Aims to eliminate domain bias in authentication without domain labels.
problem Authentication models are biased due to domain differences.
method Discover latent domains and eliminate domain difference alternately, using a meta-learning framework.
result Eliminates domain difference in authentication without domain labels.