Intrinsic Decentralized Stochastic Riemannian Optimization on Manifolds with Bounded Sectional Curvature

Duc Toan Nguyen, César A. Uribe · 2026

Read Intrinsic Decentralized Stochastic Riemannian Optimization on Manifolds with Bounded Sectional Curvature by Duc Toan Nguyen, César A. Uribe (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-17096.txt.

Description

Nguyen and Uribe study decentralized optimization on Riemannian manifolds with diminishing step sizes. Such methods matter when data are non Euclidean and spread across a network. Older intrinsic methods use constant step sizes and settle only near the optimum. The authors prove an exact bound for consensus error and an ergodic bound for the optimality gap. They call this the first exact guarantee of its kind for geodesically convex problems. They illustrate the method with distributed PCA on a Grassmann manifold.

Catalog record

CREATORDuc Toan Nguyen, César A. Uribe
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33082
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.17096

Browse the full library · About the Tower of Babel