Strong negative type in spheres

From MaRDI portal



Abstract: It is known that spheres have negative type, but only subsets with at most one pair of antipodal points have strict negative type. These are conditions on the (angular) distances within any finite subset of points. We show that subsets with at most one pair of antipodal points have strong negative type, a condition on every probability distribution of points. This implies that the function of expected distances to points determines uniquely the probability measure on such a set. It also implies that the distance covariance test for stochastic independence, introduced by Sz'ekely, Rizzo and Bakirov, is consistent against all alternatives in such sets. Similarly, it allows tests of goodness of fit, equality of distributions, and hierarchical clustering with angular distances. We prove this by showing an analogue of the Cram'er--Wold theorem.


Let \((X,d)\) be a metric space. This metric space is said to have \textit{negative type}, if for all \(n\geq 1\), all \(x_i\in X\) and \(\alpha_i\in \mathbb{R}\) for \(i=1,2,\dots,n\), the inequality \(\sum_{i=1}^n\sum_{j=1}^n \alpha_i\alpha_j d(x_i,x_j) \leq 0\). The metric space is said to have \textit{strictly negative type}, if equality holds in this inequality if and only if every \(\alpha_i\) equals zero. It is known that all Euclidean spaces have strictly negative type. A (Borel) probability measure \(\mu\) on \(X\) is said to have a \textit{finite first moment}, if for some (and hence for all) \(o\in X\), \(\int d(o,x) d\mu(x)<\infty\). Suppose that \(\mu_1,\mu_2\) have finite first moments. It can be shown that \(\int d(x_1,x_2) d(\mu_1-\mu_2)^2(x_1,x_2)\leq 0\). \((X,d)\) is said to have \textit{strong negative type}, if in this inequality equality holds if and only if \(\mu_1=\mu_2\). It is known that spheres have negative type, but only subsets with at most one pair of antipodal points have strict negative type. The paper under review shows that subsets of spheres with at most one pair of antipodal points have strong negative type. This implies that the function of expected distances to points determines uniquely the probability measure on such a set.











This page was built for publication: Strong negative type in spheres

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2206211)