DP-SIGNSGD: When Efficiency Meets Privacy and Robustness

ICASSP 2021accepted0 citations

DP-SIGNSGD: When Efficiency Meets Privacy and Robustness

Abstract

Notice of Violation of IEEE Publication Principles: "DP-SIGNSGD: When Efficiency Meets Privacy and Robustness," by Lingjuan Lyu, in the Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021, pp. 3070-3074 After careful and considered review of the content and authorship of this paper by a duly constituted expert committee, this paper has been found to be in violation of IEEE’s Publication Principles. This paper contains significant portions of content from the paper cited below that were reused with insufficient credit to the following “Stochastic-Sign SGD for Federated Learning with Theoretical Guarantees” by Richeng Jin, Yufan Huang, Xiaofan He, Huaiyu Dai, Tianfu Wu posted in ArXiv:2002.10940 Federated learning (FL) has emerged as a promising collaboration paradigm by enabling a multitude of parties to construct a joint model without exposing their private training data. Three main challenges in FL are efficiency, privacy, and robustness. The recently proposed <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SIGN</inf> SGD with majority vote shows a promising direction to deal with efficiency and Byzantine robustness. However, there is no guarantee that <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SIGN</inf> SGD is privacy-preserving. In this paper, we bridge this gap by presenting an improved method called <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DP-SIGN</inf> SGD, which can meet all the aforementioned properties. We further propose an error-feedback variant of <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DP-SIGN</inf> SGD to improve accuracy. Experimental results on benchmark image datasets demonstrate the effectiveness of our proposed methods.

BibTeX

@inproceedings{icassp2021_dpsignsgdwheneff,
  title = {DP-SIGNSGD: When Efficiency Meets Privacy and Robustness},
  author = {},
  booktitle = {ICASSP 2021},
  year = {2021}
}