Summary of Federated Source-free Domain Adaptation For Classification: Weighted Cluster Aggregation For Unlabeled Data, by Junki Mori et al.
Federated Source-free Domain Adaptation for Classification: Weighted Cluster Aggregation for Unlabeled Data
by Junki Mori, Kosuke Kihara, Taiki Miyagawa, Akinori F. Ebihara, Isamu Teranishi, Hisashi Kashima
First submitted to arxiv on: 18 Dec 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 This paper addresses a crucial challenge in Federated Learning (FL) by introducing Federated source-Free Domain Adaptation (FFREEDA), where the server has pre-trained models on labeled data from one domain, but clients only have unlabeled data from multiple target domains. The authors propose FedWCA, a novel method to mitigate both domain shifts and privacy concerns using only client-side data. FedWCA consists of three phases: private clustering for global models, weighted aggregation of these models, and local adaptation with pseudo-labeling. Experimental results demonstrate the superiority of FedWCA over existing methods in FFREEDA. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper solves a big problem in machine learning where we want to share knowledge between different devices or groups without sharing their data. It’s like trying to teach someone a new language without showing them any books or words! The authors created a new way to do this, called FedWCA, which helps the server and clients work together even when they don’t have the same kind of information. They tested it and showed that it works better than other methods in certain situations. |
Keywords
» Artificial intelligence » Clustering » Domain adaptation » Federated learning » Machine learning
