Skip to main navigation Skip to search Skip to main content

TRANSPORTATION DISTANCE BETWEEN PROBABILITY MEASURES ON THE INFINITE REGULAR TREE

  • LIGO-Massachusetts Institute of Technology
  • Tsinghua University

Research output: Contribution to journalArticlepeer-review

Abstract

In the infinite regular tree \BbbTq+1 with q \in \BbbZ\geq2, we consider families \{\munu\}, indexed by vertices u and nonnegative integers (``discrete time steps"") n, of probability measures such that \munu(v) = \munu\prime(v\prime) if the distances d(u, v) and d(u\prime, v\prime) are equal. Let d be a positive integer, and let X and Y be two vertices in the tree which are at distance d apart. We compute a formula for the transportation distance W1\bigl(\munX, \munY\bigr) in terms of generating functions. In the special case where \munu = \frakmnu are measures from simple random walks after n time steps, we establish the linear asymptotic formula W1\bigl(\frakmnX, \frakmnY\bigr) = An + B + o(1), as n \rightarrow \infty, and give the formulas for the coefficients A and B in closed forms. We also obtain linear asymptotic formulas when \munu is the uniform distribution on the sphere or on the ball of radius n as n \rightarrow \infty. We show that these six coefficients (two from the simple random walk, two from the uniform distribution on the sphere, and two from the uniform distribution on the ball) are related by inequalities.

Original languageEnglish
Pages (from-to)1113-1157
Number of pages45
JournalSIAM Journal on Discrete Mathematics
Volume38
Issue number1
DOIs
Publication statusPublished - 2024
Externally publishedYes

Keywords

  • Kantorovich problem
  • Ollivier-Ricci curvature
  • Wasserstein distance
  • asymptotic formulas
  • coarse Ricci curvature
  • generating functions
  • graph statistics
  • infinite regular tree
  • optimal transport
  • radially symmetric probability distributions
  • random walks on graphs
  • transportation distance

Fingerprint

Dive into the research topics of 'TRANSPORTATION DISTANCE BETWEEN PROBABILITY MEASURES ON THE INFINITE REGULAR TREE'. Together they form a unique fingerprint.

Cite this