Summary of Global Outlier Detection in a Federated Learning Setting with Isolation Forest, by Daniele Malpetti and Laura Azzimonti
Global Outlier Detection in a Federated Learning Setting with Isolation Forest
by Daniele Malpetti, Laura Azzimonti
First submitted to arxiv on: 20 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 A novel strategy for detecting global outliers is proposed in a federated learning setting, focusing on cross-silo scenarios. The approach involves two servers and masked local data transmission from clients to one server, preventing sensitive information disclosure while identifying outliers. To ensure privacy, a permutation mechanism is implemented, and the server uses Isolation Forest or its extended version for outlier detection, communicating results back to clients to remove outliers before federated model training. This method provides comparable results to centralized execution on plain data. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A new way to find unusual data points in a big project that shares information between different places is presented. It’s like having two secret boxes where people send their data, so nobody can see the real data but still figure out if it’s weird. The team uses special techniques to keep the data safe and then tells each person if they have any unusual data, so they can get rid of it before working together on a new project. |
Keywords
» Artificial intelligence » Federated learning » Outlier detection
