New thesis details optimization for efficient, robust federated learning

A new thesis explores theoretical foundations for optimizing machine learning models in distributed and federated settings. It introduces ProxSkip and Variance Reduced ProxSkip algorithms to improve communication efficiency and robustness, particularly under partial client participation and heterogeneous data conditions. The work also provides theoretical insights into Byzantine robustness, gradient compression techniques, and low-rank adaptation for large model fine-tuning. AI

IMPACT Provides theoretical advancements for more efficient and robust training of large-scale machine learning models.

RANK_REASON The item is a research paper detailing theoretical foundations and algorithmic improvements for distributed and federated optimization in machine learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• Connected Papers
• DagsHub
• Gotit.pub
• Hryhorii Malynovskyi
• Hugging Face
• Litmaps
• ProxSkip
• ScienceCast
• scite Smart Citations
• Variance Reduced ProxSkip

• paper
• infra

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