Summary of Corbin-fl: a Differentially Private Federated Learning Mechanism Using Common Randomness, by Hojat Allah Salehi et al.
CorBin-FL: A Differentially Private Federated Learning Mechanism using Common Randomness
by Hojat Allah Salehi, Md Jueal Mia, S. Sandeep Pradhan, M. Hadi Amini, Farhad Shirani
First submitted to arxiv on: 20 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Cryptography and Security (cs.CR); Information Theory (cs.IT)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary In this paper, researchers develop a novel approach to federated learning (FL) called CorBin-FL, which combines correlated binary stochastic quantization with secure multi-party computation to achieve differential privacy. The authors demonstrate that CorBin-FL achieves parameter-level local differential privacy and optimizes the privacy-utility trade-off between mean square error utility measure and local differential privacy. They also propose AugCorBin-FL, an extension that adds user-level and sample-level central differential privacy guarantees. Experimental results on MNIST and CIFAR10 datasets show that CorBin-FL outperforms existing differentially private FL mechanisms in terms of model accuracy under equal privacy budgets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated learning is a way for many devices to learn together without sharing their personal data. The problem is that this method can’t balance being private, efficient, and accurate at the same time. This paper introduces CorBin-FL, a new way to achieve this balance by using special math techniques. It shows that this approach keeps individual privacy safe while still making accurate predictions. The researchers also tested their idea on some famous datasets and found it works better than other methods. |
Keywords
» Artificial intelligence » Federated learning » Quantization
