Summary of Recovering Global Data Distribution Locally in Federated Learning, by Ziyu Yao
Recovering Global Data Distribution Locally in Federated Learning
by Ziyu Yao
First submitted to arxiv on: 21 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Vision and Pattern Recognition (cs.CV)
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 Federated Learning (FL) is a distributed machine learning paradigm that enables collaboration among multiple clients to train a shared model without sharing raw data. The paper proposes ReGL, a novel approach to address the challenge of label imbalance in FL. Label imbalance occurs when clients possess certain classes exclusively while having numerous minority and missing classes. ReGL recovers the global data distribution locally by using generative models to synthesize images that complement minority and missing classes. This alleviates label imbalance, making synthetic images align more closely with the global distribution. The approach is conducted at the client-side without leaking data privacy. Experiments on various image classification datasets demonstrate the superiority of ReGL over existing state-of-the-art works in tackling label imbalance in FL. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine a group of people working together to train a model, but each person has different information they’re sharing. This can cause problems if some people have more important or rare information than others. A new approach called ReGL helps solve this problem by using computer-generated images to fill in the gaps and make everything fair again. This is done without sharing any sensitive information with anyone else. By testing this approach on different datasets, researchers found that it outperforms other methods in making sure everyone’s voice is heard. |
Keywords
» Artificial intelligence » Federated learning » Image classification » Machine learning
