English

Central Limit Theorems on Compact Metric Spaces

Probability 2020-01-14 v2

Abstract

We produce a series of Central Limit Theorems (CLTs) associated to compact metric measure spaces (K,d,η)(K,d,\eta), with η\eta a reasonable probability measure. For the first CLT, we can ignore η\eta by isometrically embedding KK into C(K){\mathcal C}(K), the space of continuous functions on KK with the sup norm, and then applying known CLTs for sample means on Banach spaces (Theorem 3.1). However, the sample mean makes no sense back on KK, so using η\eta we develop a CLT for the sample Fr\'echet mean (Corollary 4.1). This involves working on the closed convex hull of the embedded image of KK. To work in the easier Hilbert space setting of L2(K,η)L^2(K,\eta), we have to modify the metric dd to a related metric dηd_\eta. We obtain an L2L^2-CLT for both the sample mean and the sample Fr\'echet mean (Theorem 5.1), and we relate the Fr\'echet sample and population means on the closed convex hull to the Fr\'echet means on the image of KK. Since the L2L^2 and LL^\infty norms play important roles, in Section 6 we develop a metric-measure criterion relating dd and η\eta under which all LpL^p norms are equivalent.

Keywords

Cite

@article{arxiv.2001.02793,
  title  = {Central Limit Theorems on Compact Metric Spaces},
  author = {Steven Rosenberg and Jie Xu},
  journal= {arXiv preprint arXiv:2001.02793},
  year   = {2020}
}
R2 v1 2026-06-23T13:06:31.531Z