Summary of Fedfisher: Leveraging Fisher Information For One-shot Federated Learning, by Divyansh Jhunjhunwala et al.
FedFisher: Leveraging Fisher Information for One-Shot Federated Learning
by Divyansh Jhunjhunwala, Shiqiang Wang, Gauri Joshi
First submitted to arxiv on: 19 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (stat.ML)
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 FedFisher is a novel algorithm for one-shot federated learning (FL), addressing the limitations of standard FL by enabling training in a single round. This Bayesian-inspired approach leverages Fisher information matrices computed on local client models to improve global model accuracy. Theoretical analysis reveals that FedFisher’s error becomes negligible as network width and local training increase. Practical variants, including diagonal Fisher and K-FAC approximations, demonstrate improved communication and compute efficiency for FL. Experimental results across various datasets show consistent performance gains over competing baselines. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper talks about a new way to train artificial intelligence models on many devices without sharing their data. This is called federated learning (FL). Traditional FL requires lots of back-and-forth communication, which can be slow and vulnerable to attacks. The authors propose a new method called FedFisher that does all the training in one go. They show that this method works better as the models get more complex and the devices do more local training. The results are promising and could lead to faster and more secure AI model development. |
Keywords
* Artificial intelligence * Federated learning * One shot
