On the Tradeoff between Privacy Preservation and Byzantine-Robustness in Decentralized Learning

1 code implementation • 28 Aug 2023 • Haoxiang Ye, Heng Zhu, Qing Ling

For a class of state-of-the-art robust aggregation rules, we give unified analysis of the "mixing abilities".

Privacy Preserving

Optimal Compression of Unit Norm Vectors in the High Distortion Regime

no code implementations • 16 Jul 2023 • Heng Zhu, Avishek Ghosh, Arya Mazumdar

We approach this problem in a worst-case scenario, without any prior information on the vector, but allowing for the use of randomized compression maps.

Bridging Differential Privacy and Byzantine-Robustness via Model Aggregation

1 code implementation • 29 Apr 2022 • Heng Zhu, Qing Ling

We analyze the trade-off between privacy preservation and learning performance, and show that the influence of our proposed DP mechanisms is decoupled with that of robust stochastic model aggregation.

Federated Learning

BROADCAST: Reducing Both Stochastic and Compression Noise to Robustify Communication-Efficient Federated Learning

1 code implementation • 14 Apr 2021 • Heng Zhu, Qing Ling

Communication between workers and the master node to collect local stochastic gradients is a key bottleneck in a large-scale federated learning system.

Federated Learning