Gaussian Cooling and O*(n^3) Algorithms for Volume and Gaussian Volume
Abstract
We present an randomized algorithm for estimating the volume of a well-rounded convex body given by a membership oracle, improving on the previous best complexity of . The new algorithmic ingredient is an accelerated cooling schedule where the rate of cooling increases with the temperature. Previously, the known approach for potentially achieving this asymptotic complexity relied on a positive resolution of the KLS hyperplane conjecture, a central open problem in convex geometry. We also obtain an randomized algorithm for integrating a standard Gaussian distribution over an arbitrary convex set containing the unit ball. Both the volume and Gaussian volume algorithms use an improved algorithm for sampling a Gaussian distribution restricted to a convex body. In this latter setting, as we show, the KLS conjecture holds and for a spherical Gaussian distribution with variance , the sampling complexity is for the first sample and for every subsequent sample.
Keywords
Cite
@article{arxiv.1409.6011,
title = {Gaussian Cooling and O*(n^3) Algorithms for Volume and Gaussian Volume},
author = {Ben Cousins and Santosh Vempala},
journal= {arXiv preprint arXiv:1409.6011},
year = {2016}
}
Comments
This paper is a combination of two previously published conference papers: "A Cubic Algorithm for Computing Gaussian Volume" (SODA 2014, arXiv:1306.5829) and "Bypassing KLS: Gaussian Cooling and an $O^*(n^3)$ Volume Algorithm" (STOC 2015). Additionally, this version has a major simplification to the main proof in the latter conference paper. (Lemma 3.2 in this version) 36 pages