A note on estimating the dimension from a random geometric graph
Abstract
Let be a random geometric graph with vertex set based on i.i.d.\ random vectors drawn from an unknown density on . An edge is present when , for a given threshold possibly depending upon , where denotes Euclidean distance. We study the problem of estimating the dimension of the underlying space when we have access to the adjacency matrix of the graph but do not know or the vectors . The main result of the paper is that there exists an estimator of that converges to in probability as for all densities with whenever and . The conditions allow very sparse graphs since when , the graph contains isolated edges only, with high probability. We also show that, without any condition on the density, a consistent estimator of exists when and .
Cite
@article{arxiv.2311.13059,
title = {A note on estimating the dimension from a random geometric graph},
author = {Caelan Atamanchuk and Luc Devroye and Gabor Lugosi},
journal= {arXiv preprint arXiv:2311.13059},
year = {2023}
}