Summary of Federated Optimization with Doubly Regularized Drift Correction, by Xiaowen Jiang et al.
Federated Optimization with Doubly Regularized Drift Correction
by Xiaowen Jiang, Anton Rodomanov, Sebastian U. Stich
First submitted to arxiv on: 12 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Optimization and Control (math.OC)
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 This paper presents a critical examination of federated learning, a distributed optimization approach that enables training machine learning models on decentralized devices while keeping data localized. The authors identify the limitations of the standard FedAvg method, which is prone to client drift, a phenomenon that can lead to decreased performance and increased communication costs compared to centralized methods. To address this issue, previous works have proposed various strategies to mitigate drift, but none have consistently demonstrated improved communication-computation trade-offs over vanilla gradient descent. |
Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated learning is a way for lots of devices to work together to train machine learning models without sharing their data with each other. The problem is that sometimes these devices can start to learn different things than they were supposed to, which makes it harder to get good results and takes more effort to communicate. Some people have tried to fix this by coming up with new ways to handle the differences between devices, but so far none of those solutions have been clearly better than just using a simple training method. |
Keywords
» Artificial intelligence » Federated learning » Gradient descent » Machine learning » Optimization