View Related Documents

Abstract

Denote by KnK_n the convex hull of nn independent random points distributed uniformly in a convex body KK in \Rd\R^d, by VnV_n the volume of KnK_n, by DnD_n the volume of K\KnK\backslash K_n, and by NnN_n the number of vertices of KnK_n. A well-known identity due to Efron relates the expected volume EDn{\it ED}_n---and thus EVn{\it EV}_n---to the expected number ENn+1{\it EN}_{n+1}. This identity is extended from expected values to higher moments. The planar case of the arising identity for the variances provides in a simple way the corrected version of a central limit theorem for DnD_n by Cabo and Groeneboom (KK being a convex polygon) and an improvement of a central limit theorem for DnD_n by Hsing (KK being a circular disk). Estimates of \var Dn\var D_n (KK being a two-dimensional smooth convex body) and \var Nn\var N_n (KK being a dd-dimensional smooth convex body, d ³ 4d\geq 4) are obtained. The identity for moments of arbitrary order shows that the distribution of NnN_n determines EVn-1, EVn-22,..., EVd+1n-d-1{\it EV}_{n-1}, {\it EV}_{n-2}^2,\dots, {\it EV}_{d+1}^{n-d-1}. Reversely it is proved that these n-d-1n-d-1 moments determine the distribution of NnN_n entirely. The resulting formula for the probability that Nn=k (k=d+1,... , n)N_n=k\ (k=d+1,\dots , n) appears to be new for k ³ d+2k\geq d+2 and yields an answer to a question raised by Baryshnikov. For k=d+1k=d+1 the formula reduces to an identity which has been repeatedly pointed out.

Fulltext Preview

Image of the first page of the fulltext document