Summary of Semi-variance Reduction For Fair Federated Learning, by Saber Malekmohammadi
Semi-Variance Reduction for Fair Federated Learning
by Saber Malekmohammadi
First submitted to arxiv on: 23 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computers and Society (cs.CY)
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 research aims to develop fair Federated Learning (FL) algorithms that balance individual client performance with the overall system’s average performance. Existing fair FL methods prioritize the worst-performing clients, often at the expense of well-performing ones. Inspired by risk modeling techniques in Finance, two new algorithms, Variance Reduction (VRed) and Semi-Variance Reduction (SemiVRed), are introduced to promote equality or focus on improving the worst-off clients’ performance. The proposed methods, VRed and SemiVRed, are evaluated through extensive experiments on various vision and language datasets, demonstrating that SemiVRed achieves state-of-the-art performance in scenarios with heterogeneous data distributions while maintaining fairness and improving overall system performance. |
Low | GrooveSquid.com (original content) | Low Difficulty Summary The research aims to make a Federated Learning system fair for all participating clients. Right now, most algorithms prioritize the worst-performing clients, which can hurt well-performing ones. The scientists came up with two new ideas, Variance Reduction (VRed) and Semi-Variance Reduction (SemiVRed), inspired by financial risk models. They tested these methods on many datasets and found that one of them, SemiVRed, does a great job in making the system fair and good overall. |
Keywords
» Artificial intelligence » Federated learning