New FL-OA framework enhances federated learning against Byzantine attacks

Researchers have introduced FL-OA, a novel federated learning framework designed to enhance robustness against Byzantine attacks. This framework utilizes outsourced auditing with a third-party organization that possesses a root dataset, allowing for robust aggregation without stringent assumptions about the proportion of malicious devices. FL-OA also incorporates a gradient ascent step and a correction term to address divergence in benign updates and employs a parameter importance indicator to simplify the auditing process by focusing on critical parameters. AI

IMPACT This framework could improve the security and reliability of collaborative AI model training across distributed devices.

RANK_REASON The cluster contains an academic paper detailing a new framework for federated learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Byzantine attacks
• Federated learning
• FL-OA

• paper
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →