English

Finite Volume Spaces and Sparsification

Data Structures and Algorithms 2010-08-03 v3 Discrete Mathematics

Abstract

We introduce and study finite dd-volumes - the high dimensional generalization of finite metric spaces. Having developed a suitable combinatorial machinery, we define 1\ell_1-volumes and show that they contain Euclidean volumes and hypertree volumes. We show that they can approximate any dd-volume with O(nd)O(n^d) multiplicative distortion. On the other hand, contrary to Bourgain's theorem for d=1d=1, there exists a 22-volume that on nn vertices that cannot be approximated by any 1\ell_1-volume with distortion smaller than Ω~(n1/5)\tilde{\Omega}(n^{1/5}). We further address the problem of 1\ell_1-dimension reduction in the context of 1\ell_1 volumes, and show that this phenomenon does occur, although not to the same striking degree as it does for Euclidean metrics and volumes. In particular, we show that any 1\ell_1 metric on nn points can be (1+ϵ)(1+ \epsilon)-approximated by a sum of O(n/ϵ2)O(n/\epsilon^2) cut metrics, improving over the best previously known bound of O(nlogn)O(n \log n) due to Schechtman. In order to deal with dimension reduction, we extend the techniques and ideas introduced by Karger and Bencz{\'u}r, and Spielman et al.~in the context of graph Sparsification, and develop general methods with a wide range of applications.

Keywords

Cite

@article{arxiv.1002.3541,
  title  = {Finite Volume Spaces and Sparsification},
  author = {Ilan Newman and Yuri Rabinovich},
  journal= {arXiv preprint arXiv:1002.3541},
  year   = {2010}
}

Comments

previous revision was the wrong file: the new revision: changed (extended considerably) the treatment of finite volumes (see revised abstract). Inserted new applications for the sparsification techniques

R2 v1 2026-06-21T14:48:32.257Z