Random inscribed polytopes in projective geometries (Q832480): Difference between revisions

This work deals with central limit theorems for random inscribed polytopes. The main result is based on the following data: a convex body \(K\) of class \(C^2_+\) in \({\mathbb R}^d\), a probability measure \(\sigma\) on \(\operatorname{bd}K\) with a continuous, positive density, a measure \(\Phi\) on \(K\) with a positive density which is continuous on a neighborhood of \(\operatorname{bd}K\) in \(K\). Let \(K_\sigma(n)\) be the convex hull of \(n\) independent random points on \(\operatorname{bd}K\), each with distribution \(\sigma\). A central limit theorem, as \(n\to\infty\), (in fact, a quantitative improvement) is proved for the \(\Phi\)-measure of \(K_\sigma(n)\). This is a counterpart to an earlier result of \textit{F. Besau} and \textit{C. Thäle} [Trans. Am. Math. Soc. 373, No. 12, 8911--8941 (2020; Zbl 1468.52006)], where the random points are inside \(K\). From the main result, several central limit theorems are deduced, in spherical and hyperbolic geometry and in projective Finsler spaces, but also for dual volumes and for the mean width of random polyhedral sets circumscribing a convex body. The proof of the main result makes use of weighted surface bodies and of a normal approximation bound due to \textit{R. Lachièze-Rey} and \textit{G. Peccati} [Ann. Appl. Probab. 27, No. 4, 1992--2031 (2017; Zbl 1374.60023)]. A further important ingredient is a lower bound for the variance of \(\Phi(K_\sigma(n))\).