Sign Up to like & get
recommendations!
0
Published in 2024 at "Electronic Journal of Probability"
DOI: 10.1214/25-ejp1418
Abstract: Hit-and-Run is a coordinate-free Gibbs sampler, yet the quantitative advantages of its coordinate-free property remain largely unexplored beyond empirical studies. In this paper, we prove sharp estimates for the Wasserstein contraction of Hit-and-Run in Gaussian…
read more here.
Keywords:
randomized kaczmarz;
accelerated convergence;
kaczmarz algorithm;
coordinate free ... See more keywords