Summary of Ross:robust Decentralized Stochastic Learning Based on Shapley Values, by Lina Wang et al.
ROSS:RObust decentralized Stochastic learning based on Shapley values
by Lina Wang, Yunsheng Yuan, Feng Li, Lingjie Duan
First submitted to arxiv on: 1 Nov 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 The proposed ROSS algorithm is a novel decentralized stochastic learning method that tackles heterogeneity in distributed datasets without relying on central servers. The approach aggregates cross-gradient information from neighbors to update local models, with Shapley values used to weight these contributions. Theoretical analysis shows linear convergence speedup, and experimental results on public datasets demonstrate improved performance over state-of-the-art methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary ROSS is a new way for agents to work together and learn from each other without needing a central server. It helps solve problems when the data is not the same across all the agents. The algorithm looks at how the models are changing compared to its neighbors’ data, and uses something called Shapley values to decide which parts of that information to use. This helps it work better in situations where some data might be bad or different from the rest. |
