New personalized federated learning method HDC-FL shows strong performance

Researchers have introduced HDC-FL, a novel method for personalized federated learning that leverages client-specific classifier heads to guide representation sharing. This approach groups client descriptors and creates encoder prototypes for each group, which are then returned to clients. HDC-FL demonstrates competitive performance against existing personalized and clustered federated learning baselines across various CIFAR-10 and CIFAR-100 configurations, particularly excelling in macro-averaged worst-client and bottom-20% accuracy. AI

IMPACT This method could improve the efficiency and effectiveness of federated learning systems by better managing client representation sharing.

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

Read on arXiv cs.LG →

• Amirardalan Dehghanpour
• arXiv
• CIFAR-10
• CIFAR-100
• HDC-FL
• MobileNetV2
• ResNet-18

• paper
• other

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