paper-with-me

홈 › Papers

On Mitigating the Utility-Loss in Differentially Private Learning: A new Perspective by a Geometrically Inspired Kernel Approach

2023-04-03 · Mohit Kumar, Bernhard A. Moser, Lukas Fischer

Privacy-utility tradeoff remains as one of the fundamental issues of differentially private machine learning. This paper introduces a geometrically inspired kernel-based approach to mitigate the accuracy-loss issue in classification. In this approach, a representation of the affine hull of given data points is learned in Reproducing Kernel Hilbert Spaces (RKHS). This leads to a novel distance measure that hides privacy-sensitive information about individual data points and improves the privacy-utility tradeoff via significantly reducing the risk of membership inference attacks. The effectiveness of the approach is demonstrated through experiments on MNIST dataset, Freiburg groceries dataset, and a real biomedical dataset. It is verified that the approach remains computationally practical. The application of the approach to federated learning is considered and it is observed that the accuracy-loss due to data being distributed is either marginal or not significantly high.

📄 PDF Abstract BibTeX arXiv:2304.01300

Code (0)

등록된 구현이 없습니다.

Tasks

Federated LearningPrivacy PreservingRepresentation Learning

Similar Papers 제목 키워드 기반

DPlis: Boosting Utility of Differentially Private Deep Learning via Randomized Smoothing

2021-03-02 · Wenxiao Wang, Tianhao Wang, Lun Wang, Nanqing Luo 외

Deep learning techniques have achieved remarkable performance in wide-ranging tasks. However, when trained on privacy-sensitive datasets, the model parameters may expose private information in training data. Prior attemp…

The role of self-supervised pretraining in differentially private medical image analysis

2026-01-27 · Soroosh Tayebi Arasteh, Mina Farajiamiri, Mahshad Lotfinia, Behrus Hinrichs-Puladi 외 arxiv

Differential privacy (DP) provides formal protection for sensitive data but typically incurs substantial losses in diagnostic performance. Model initialization has emerged as a critical factor in mitigating this degradat…

Self-Supervised Learning

Mitigating Statistical Bias within Differentially Private Synthetic Data

2021-08-24 · Sahra Ghalebikesabi, Harrison Wilde, Jack Jewson, Arnaud Doucet 외

Increasing interest in privacy-preserving machine learning has led to new and evolved approaches for generating private synthetic data from undisclosed real data. However, mechanisms of privacy preservation can significa…

Privacy Preserving

DPAF: Image Synthesis via Differentially Private Aggregation in Forward Phase

2023-04-20 · Chih-Hsun Lin, Chia-Yi Hsu, Chia-Mu Yu, Yang Cao 외

Differentially private synthetic data is a promising alternative for sensitive data release. Many differentially private generative models have been proposed in the literature. Unfortunately, they all suffer from the low…

Image Generation

An end-to-end Differentially Private Latent Dirichlet Allocation Using a Spectral Algorithm

2018-05-25 · ICML 2020 1 · Christopher DeCarolis, Mukul Ram, Seyed A. Esmaeili, Yu-Xiang Wang 외

We provide an end-to-end differentially private spectral algorithm for learning LDA, based on matrix/tensor decompositions, and establish theoretical guarantees on utility/consistency of the estimated model parameters. T…

SensitivityVariational Inference