arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Author

Benjamin Capdeville

2 papers · 2024–2026

2024: 12025: 02026: 1
2024peak 1 in 20242026

Frequent co-authors

Papers

showing 1–2 of 2 RSS

Isometric embedding of the n-point spaces into the space of spaces for $n \leq 4$

Benjamin Capdeville

math.MG · · cited 0× · arXiv ↗ · more like this ·

In [The Space of Spaces: Curvature Bounds and Gradient Flows on the Space of Metric Measure Spaces. Memoirs of the American Mathematical Society. American Mathematical Society, 2023], Sturm studied the space of all metric measure spaces up to isomorphism which he called The space of spaces. He also introduced for a natural number n the space of all n-points metric spaces. The aim of this article is to study if the embedding of this space in the space of spaces is isometric. Using results from [Haggai Maron and Yaron Lipman. (probably) concave graph matching. Advances in Neural Information Processing Systems, 31, 2018] and [Hiroshi Maehara. Euclidean embeddings of finite metric spaces. Discrete Mathematics, 2013], we prove that it is the case for $n \leq 4$ and for Euclidean metric spaces.

Mirror Langevin diffusions: Convergence rates and Markov chain approximations

Benjamin Capdeville, Young-Heon Kim, Soumik Pal

math.PR · · arXiv ↗ · more like this ·

Given a strongly convex function $u$, equip $R^d$ with a Riemannian metric given by the Hessian $\nabla^2 u$. This is a so-called Hessian manifold. Given a probability density $μ$ one may run a Langevin diffusion intrinsic to the manifold with stationary distribution $μ$. Such (Hessian) manifold-valued Langevin diffusions are called Mirror Langevin diffusions (MLD) which have recently become popular. One of the questions we explore is whether, given $μ$, one can choose $u$ to get an exponential convergence to equilibrium for the MLD, especially if $μ$ is not strongly log-concave. Our results are based on Lyapunov function methods and give sufficient conditions for a Poincaré or a log-Sobolev inequality to hold for the MLD. These, in turn, imply exponential convergence. We also introduce a Markov chain approximation to the MLD given by a two step Gibbs sampler with stationary…