On the isotropic constant of random polytopes (Q5890511)

A convex body \(K \subset \mathbb{R}^n\) is called isotropic if it has volume \(|K|=1\), its center of mass is at \(0\) (i.e., \(\int_K x dx = 0\)) and for every \(\theta \in \mathbb{S}^{n-1}\) one has \[ \int_K \langle x, \theta \rangle dx = L_K^2, \] in which \(L_K\) is a constant independent of \(\theta\). The constant \(L_K\) is then called the isotropic constant of \(K\). Since every convex body has a unique affine image that is isotropic, one defines the isotropic constant of an arbitrary convex body as the isotropic constant of its isotropic image. The aim of the paper under review is to estimate the isotropic constant of particular random polytopes; i.e. let \(X_1, \ldots, X_N\) be independent random vectors uniformly distributed on an isotropic convex body \(K \subset \mathbb{R}^n \) and let \(K_N\) be the symmetric convex hull of the \(X_i\)'s. Then, the authors show (Theorem 1.1) that with high probability \[ L_{K_N} \leq C \sqrt{\log(2N/n) }, \] where \(C\) is an absolute constant. In Theorem 1.2 they extend their result to the symmetric convex hulls of vectors \(y_1 X_1, \ldots, y_N X_N\) with \(y=(y_1, \ldots, y_N) \in \mathbb{R}^n\) and discuss the case of a random vector \(y\). The authors note at the end of the paper that their Theorem 1.1 has subsequently been proved in [\textit{A. Giannopoulos} et al., Adv. Appl. Math. 75, 116--143 (2016; Zbl 1336.52005)] using essentially the same approach.