Summary of Learn What You Need in Personalized Federated Learning, by Kexin Lv et al.
Learn What You Need in Personalized Federated Learning
by Kexin Lv, Rui Ye, Xiaolin Huang, Jie Yang, Siheng Chen
First submitted to arxiv on: 16 Jan 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 In this paper, researchers tackle the issue of data heterogeneity in federated learning by proposing a novel algorithm-unrolling-based framework called Learn2pFed. This approach enables each local client to adaptively select which part of its model parameters should participate in collaborative training, unlike previous methods that blindly incorporate either full or partial parameters. The key innovation is optimizing the participation degree of local model parameters as learnable parameters via algorithm unrolling methods. This leads to two benefits: mathematically determining participation and obtaining more stable solutions. Experimental results on various tasks show Learn2pFed outperforms previous personalized federated learning methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Personalized federated learning wants to solve the problem of different data across local devices. Right now, methods either use all or some parts of model parameters without thinking about each device’s data. This makes bad results. To fix this, researchers created Learn2pFed, a new way for devices to work together and pick which parts of their models to share. This helps find the best combination of shared and local learning. It works better than other methods and is tested on different tasks like image recognition. |
Keywords
* Artificial intelligence * Federated learning
