English

A method to construct exponential families by representation theory

Statistics Theory 2022-10-14 v4 Machine Learning Probability Statistics Theory

Abstract

In this paper, we give a method to construct "good" exponential families systematically by representation theory. More precisely, we consider a homogeneous space G/HG/H as a sample space and construct an exponential family invariant under the transformation group GG by using a representation of GG. The method generates widely used exponential families such as normal, gamma, Bernoulli, categorical, Wishart, von Mises, Fisher-Bingham and hyperboloid distributions.

Keywords

Cite

@article{arxiv.1811.01394,
  title  = {A method to construct exponential families by representation theory},
  author = {Koichi Tojo and Taro Yoshino},
  journal= {arXiv preprint arXiv:1811.01394},
  year   = {2022}
}
R2 v1 2026-06-23T05:03:32.792Z