Subsampling is not Magic: Why Large Batch Sizes Work for Differentially Private Stochastic Optimisation

6 February 2024

Papers citing "Subsampling is not Magic: Why Large Batch Sizes Work for Differentially Private Stochastic Optimisation"

11 / 11 papers shown

Title
Avoiding Pitfalls for Privacy Accounting of Subsampled Mechanisms under Composition C. Lebeda Matthew Regehr Gautam Kamath Thomas Steinke 97 10 0 27 May 2024
Automatic Clipping: Differentially Private Deep Learning Made Easier and Stronger Zhiqi Bu Yu Wang Sheng Zha George Karypis 103 71 0 14 Jun 2022
Optimal Accounting of Differential Privacy via Characteristic Function Yuqing Zhu Jinshuo Dong Yu Wang 52 102 0 16 Jun 2021
Numerical Composition of Differential Privacy Sivakanth Gopi Y. Lee Lukas Wutschitz 54 179 0 05 Jun 2021
Privacy-preserving Data Sharing on Vertically Partitioned Data Razane Tajeddine Hibiki Ito Samuel Kaski Antti Honkela FedML 50 8 0 19 Oct 2020
Tight Differential Privacy for Discrete-Valued Mechanisms and for the Subsampled Gaussian Mechanism Using FFT A. Koskela Hibiki Ito Lukas Prediger Antti Honkela 42 59 0 12 Jun 2020
Rényi Differential Privacy of the Sampled Gaussian Mechanism Ilya Mironov Kunal Talwar Li Zhang 73 283 0 28 Aug 2019
Improving the Gaussian Mechanism for Differential Privacy: Analytical Calibration and Optimal Denoising Borja Balle Yu Wang MLT 66 403 0 16 May 2018
On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima N. Keskar Dheevatsa Mudigere J. Nocedal M. Smelyanskiy P. T. P. Tang ODL 417 2,935 0 15 Sep 2016
Deep Learning with Differential Privacy Martín Abadi Andy Chu Ian Goodfellow H. B. McMahan Ilya Mironov Kunal Talwar Li Zhang FedML SyDa 191 6,109 0 01 Jul 2016
On Sampling, Anonymization, and Differential Privacy: Or, k-Anonymization Meets Differential Privacy Ninghui Li Wahbeh H. Qardaji D. Su 100 279 0 13 Jan 2011