English

Bi-invariant Dissimilarity Measures for Sample Distributions in Lie Groups

Methodology 2024-02-21 v1 Statistics Theory Statistics Theory

Abstract

Data sets sampled in Lie groups are widespread, and as with multivariate data, it is important for many applications to assess the differences between the sets in terms of their distributions. Indices for this task are usually derived by considering the Lie group as a Riemannian manifold. Then, however, compatibility with the group operation is guaranteed only if a bi-invariant metric exists, which is not the case for most non-compact and non-commutative groups. We show here that if one considers an affine connection structure instead, one obtains bi-invariant generalizations of well-known dissimilarity measures: a Hotelling T2T^2 statistic, Bhattacharyya distance and Hellinger distance. Each of the dissimilarity measures matches its multivariate counterpart for Euclidean data and is translation-invariant, so that biases, e.g., through an arbitrary choice of reference, are avoided. We further derive non-parametric two-sample tests that are bi-invariant and consistent. We demonstrate the potential of these dissimilarity measures by performing group tests on data of knee configurations and epidemiological shape data. Significant differences are revealed in both cases.

Keywords

Cite

@article{arxiv.2402.12901,
  title  = {Bi-invariant Dissimilarity Measures for Sample Distributions in Lie Groups},
  author = {Martin Hanik and Hans-Christian Hege and Christoph von Tycowicz},
  journal= {arXiv preprint arXiv:2402.12901},
  year   = {2024}
}

Comments

An incomplete (and thus incorrect) statement in the background section on the connection of the CCS connection to Riemannian metrics was corrected. It was not used anywhere in the paper

R2 v1 2026-06-28T14:54:20.142Z