English

Metric random matchings with applications

Data Structures and Algorithms 2017-03-27 v1

Abstract

Let ({1,2,,n},d)(\{1,2,\ldots,n\},d) be a metric space. We analyze the expected value and the variance of i=1n/2d(π(2i1),π(2i))\sum_{i=1}^{\lfloor n/2\rfloor}\,d({\boldsymbol{\pi}}(2i-1),{\boldsymbol{\pi}}(2i)) for a uniformly random permutation π{\boldsymbol{\pi}} of {1,2,,n}\{1,2,\ldots,n\}, leading to the following results: (I) Consider the problem of finding a point in {1,2,,n}\{1,2,\ldots,n\} with the minimum sum of distances to all points. We show that this problem has a randomized algorithm that (1) always outputs a (2+ϵ)(2+\epsilon)-approximate solution in expected O(n/ϵ2)O(n/\epsilon^2) time and that (2) inherits Indyk's~\cite{Ind99, Ind00} algorithm to output a (1+ϵ)(1+\epsilon)-approximate solution in O(n/ϵ2)O(n/\epsilon^2) time with probability Ω(1)\Omega(1), where ϵ(0,1)\epsilon\in(0,1). (II) The average distance in ({1,2,,n},d)(\{1,2,\ldots,n\},d) can be approximated in O(n/ϵ)O(n/\epsilon) time to within a multiplicative factor in [1/2ϵ,1][\,1/2-\epsilon,1\,] with probability 1/2+Ω(1)1/2+\Omega(1), where ϵ>0\epsilon>0. (III) Assume dd to be a graph metric. Then the average distance in ({1,2,,n},d)(\{1,2,\ldots,n\},d) can be approximated in O(n)O(n) time to within a multiplicative factor in [1ϵ,1+ϵ][\,1-\epsilon,1+\epsilon\,] with probability 1/2+Ω(1)1/2+\Omega(1), where ϵ=ω(1/n1/4)\epsilon=\omega(1/n^{1/4}).

Keywords

Cite

@article{arxiv.1703.08433,
  title  = {Metric random matchings with applications},
  author = {Ching-Lueh Chang},
  journal= {arXiv preprint arXiv:1703.08433},
  year   = {2017}
}

Comments

arXiv admin note: substantial text overlap with arXiv:1702.03106

R2 v1 2026-06-22T18:55:58.824Z