Summary of Incentivizing Truthful Collaboration in Heterogeneous Federated Learning, by Dimitar Chakarov et al.
Incentivizing Truthful Collaboration in Heterogeneous Federated Learning
by Dimitar Chakarov, Nikita Tsoy, Kristian Minchev, Nikola Konstantinov
First submitted to arxiv on: 1 Dec 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Science and Game Theory (cs.GT); 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 Federated learning (FL) is a collaborative method where clients share gradient updates to learn together. However, FL is vulnerable to manipulated updates from clients. This paper explores the impact of data heterogeneity on clients’ incentives to manipulate their updates. The authors show that manipulations can lead to diminishing model performance and formulate a game to prevent such modifications. They develop a payment rule that disincentivizes sending modified updates under the FedSGD protocol, providing explicit bounds on payments and convergence rate. An experimental evaluation of the scheme is conducted in three tasks in computer vision and natural language processing, demonstrating successful disincentivization of modifications. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how people share information to learn together without sharing their raw data. They found that some people might try to cheat by changing what they send to get a better outcome. The authors created a way to make it not worth cheating, so everyone works together honestly. They tested this idea on three different tasks and showed that it worked well. |
Keywords
» Artificial intelligence » Federated learning » Natural language processing
