Summary of Ferrari: Federated Feature Unlearning Via Optimizing Feature Sensitivity, by Hanlin Gu et al.
Ferrari: Federated Feature Unlearning via Optimizing Feature Sensitivity
by Hanlin Gu, Win Kent Ong, Chee Seng Chan, Lixin Fan
First submitted to arxiv on: 23 May 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 This paper addresses the growing need for Federated Unlearning (FU) in machine learning, particularly in Federated Learning (FL). The authors highlight the importance of allowing clients to request data deletion from model service providers. They propose a feature unlearning framework called Ferrari, which minimizes feature sensitivity using Lipschitz continuity as an evaluation metric. This framework is designed for FL and aims to address limitations in existing methods that require participation from multiple clients. Experimental results demonstrate the effectiveness of Ferrari across various scenarios, including sensitive, backdoor, and biased features. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated learning lets machines learn together without sharing all their data. That’s great, but it also means we need a way for individual machines to ask for their old data back – like deleting old files from the cloud. This is called “unlearning.” The problem is that current methods need help from lots of other machines to work. Researchers have come up with a new way to unlearn features without needing so much help. They call it Ferrari, and it’s better at deleting sensitive information than previous methods. |
Keywords
» Artificial intelligence » Federated learning » Machine learning
