New decentralized optimization method achieves optimal regret bound

Researchers have developed a new decentralized online optimization method for strongly geodesically convex functions on Riemannian manifolds. This advancement addresses the unexplored strongly g-convex regime in decentralized settings, unlike previous methods that only handled g-convex losses. The new approach achieves an optimal $O(\log T)$ static regret bound, matching the rate for strongly-convex Euclidean online optimization, and also extends this bound to the two-point bandit feedback setting. AI

RANK_REASON The cluster contains a research paper detailing a new optimization method. [lever_c_demoted from research: ic=1 ai=0.4]

Read on arXiv cs.LG →

• arXiv
• Decentralized Online Riemannian Gradient Descent
• Decentralized Online Riemannian Optimization
• Decentralized Riemannian Optimization
• Euclidean online optimization
• Strongly Geodesically Convex
• Two-Point Bandit Feedback

• paper
• other

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