Summary of Foogd: Federated Collaboration For Both Out-of-distribution Generalization and Detection, by Xinting Liao et al.
FOOGD: Federated Collaboration for Both Out-of-distribution Generalization and Detection
by Xinting Liao, Weiming Liu, Pengyang Zhou, Fengyuan Yu, Jiahe Xu, Jun Wang, Wenjie Wang, Chaochao Chen, Xiaolin Zheng
First submitted to arxiv on: 15 Oct 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 paper proposes FOOGD, a federated learning (FL) method that addresses the simultaneous occurrence of various out-of-distribution (OOD) shifts in real-world scenarios. Current FL research typically focuses on either covariate-shift or semantic-shift data, overlooking their coexistence. FOOGD estimates the probability density of each client and provides reliable global distribution as guidance for the subsequent FL process. The method consists of two components: SM3D, which detects semantic-shift data powerfully, and SAG, which provides invariant yet diverse knowledge for both local covariate-shift generalization and client performance generalization. FOOGD significantly outperforms existing methods in empirical validations by reliably estimating non-normalized decentralized distributions, detecting semantic shift data via score values, and generalizing to covariate-shift data by regularizing feature extractor. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine you’re trying to learn something new with the help of other people who have different ways of understanding it. This is like a big puzzle that needs to be solved together. But sometimes, some pieces don’t fit or are missing, making it harder to solve. The researchers in this paper came up with a way to make sure all the pieces fit together and you can learn from each other even when things get complicated. They created a method called FOOGD that helps people learn new things by working together. It’s like having a special guide that shows you how to put all the puzzle pieces together correctly. |
Keywords
» Artificial intelligence » Federated learning » Generalization » Probability
