Summary of Mitigating Malicious Attacks in Federated Learning Via Confidence-aware Defense, by Qilei Li et al.
Mitigating Malicious Attacks in Federated Learning via Confidence-aware Defense
by Qilei Li, Ahmed M. Abdelmoniem
First submitted to arxiv on: 5 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Distributed, Parallel, and Cluster Computing (cs.DC)
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 Federated Learning framework called Confidence-Aware Defense (CAD) is proposed to detect and mitigate malicious updates from attacked clients, addressing both data poisoning and model poisoning attacks. The CAD method utilizes confidence scores of local models as a criterion to evaluate the reliability of updates. It’s based on the observation that malicious attacks cause increased uncertainty in predictions, deviating from the previous state. This comprehensive defense mechanism effectively identifies and addresses potential malicious updates, enhancing the robustness of FL systems against various attack types and scenarios. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated Learning (FL) is a way for many devices to work together on a big project without sharing their own information. But sometimes, this system can be tricked by bad actors who try to ruin the results. To fix this, scientists created a new method called Confidence-Aware Defense (CAD). It checks how sure each device is about its answers and ignores any that are suspicious. This helps keep the final result accurate and good. |
Keywords
* Artificial intelligence * Federated learning
