Summary of Contribution Evaluation Of Heterogeneous Participants in Federated Learning Via Prototypical Representations, by Qi Guo et al.
Contribution Evaluation of Heterogeneous Participants in Federated Learning via Prototypical Representations
by Qi Guo, Minghao Yao, Zhen Tian, Saiyu Qi, Yong Qi, Yun Lin, Jin Song Dong
First submitted to arxiv on: 2 Jul 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 Medium Difficulty summary: This paper explores novel methods for evaluating contributions in federated learning (FL), a crucial research area with applications in detecting low-quality datasets, enhancing model robustness, and designing incentive mechanisms. Existing approaches rely on data volume, model similarity, and auxiliary test datasets but struggle with heterogeneous data distributions. The proposed method, FLCE, introduces a new indicator called class contribution momentum to refine contribution evaluation. This approach considers individual, relative, and holistic perspectives, eliminating the need for an auxiliary test dataset. Experimental results demonstrate the superiority of FLCE in terms of fidelity, effectiveness, efficiency, and heterogeneity across various scenarios. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Low Difficulty summary: Imagine a way to figure out how much each participant is contributing to a shared learning project. This paper proposes a new method to do just that in a type of collaborative learning called federated learning. The old ways of doing this relied too heavily on the amount of data or the similarity between models, which didn’t work well when the data was different. The new approach, FLCE, looks at individual contributions and how they fit together as a whole. This method is better than the old ones because it doesn’t need extra test data to work well. |
Keywords
* Artificial intelligence * Federated learning
