Summary of Collaboratively Learning Federated Models From Noisy Decentralized Data, by Haoyuan Li et al.
Collaboratively Learning Federated Models from Noisy Decentralized Data
by Haoyuan Li, Mathias Funk, Nezihe Merve Gürel, Aaqib Saeed
First submitted to arxiv on: 3 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 The proposed Federated Noise-Sifting (FedNS) approach is a straightforward yet effective method for identifying clients with low-quality data at the initial stage of Federated Learning (FL). This is particularly important as local data can be susceptible to corruption by various forms of noise and perturbations, compromising the aggregation process and leading to a subpar global model. FedNS integrates with existing FL strategies and enhances the global model’s performance by up to 13.68% in IID settings and 15.85% in non-IID settings when learning from noisy decentralized data. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary In this paper, researchers tackle the problem of noisy data in Federated Learning. They propose a new approach called Federated Noise-Sifting (FedNS) that helps identify clients with low-quality data. This is important because local data can be corrupted by noise and perturbations, making it hard to get a good global model. The authors test their method on different datasets and show that it can improve the performance of the global model by up to 15%. |
Keywords
» Artificial intelligence » Federated learning
