Abstract:We study various information-theoretic measures and the information geometry of the Poincaré distributions and the related hyperboloid distributions, and prove that their statistical mixture models are universal density estimators of smooth densities in hyperbolic spaces. The Poincaré and the hyperboloid distributions are two types of hyperbolic probability distributions defined using different models of hyperbolic geometry. Namely, the Poincaré distributions form a triparametric bivariate exponential family whose sample space is the hyperbolic Poincaré upper-half plane and natural parameter space is the open 3D convex cone of two-by-two positive-definite matrices. The family of hyperboloid distributions form another exponential family which has sample space the forward sheet of the two-sheeted unit hyperboloid modeling hyperbolic geometry. In the first part, we prove that all $f$-divergences between Poincaré distributions can be expressed using three canonical terms using Eaton's framework of maximal group invariance. We also show that the $f$-divergences between any two Poincaré distributions are asymmetric except when those distributions belong to a same leaf of a particular foliation of the parameter space. We report closed-form formula for the Fisher information matrix, the Shannon's differential entropy and the Kullback-Leibler divergence. and Bhattacharyya distances between such distributions using the framework of exponential families. In the second part, we state the corresponding results for the exponential family of hyperboloid distributions by highlighting a parameter correspondence between the Poincaré and the hyperboloid distributions. Finally, we describe a random generator to draw variates and present two Monte Carlo methods to stochastically estimate numerically $f$-divergences between hyperbolic distributions.

Information geometry and $α$-parallel prior of the beta-logistic distribution

The Asymptotic Distribution and Berry–Esseen Bound of a New Test for Independence in High Dimension with an Application to Stochastic Optimization

On Beta-Product Convolutions

Weyl Prior and Bayesian Statistics

The Alpha-Beta-Skew-Logistic Distribution And Its Applications

The Representative Points of Generalized Alpha Skew-t Distribution and Applications

Bernstein von-Mises Theorem for g-prior and nonlocal prior

Simple Alpha-Mu Approximation To Lognormal Sums

Information measures and geometry of the hyperbolic exponential families of Poincaré and hyperboloid distributions

Some Relationships and Properties of the Hypergeometric Distribution

The Generalized-Alpha-Beta-Skew-Normal Distribution: Properties and Applications

On Ordered Beta Distribution and the Generalized Incomplete Beta Function

On the Generalized Arcsine Probability Distribution with Bounded Support

A Generalized-Alpha–Beta-Skew Normal Distribution with Applications

Generalized Log-Lindley Distribution and Its Applications in Stochastic Comparison

Distributions on symmetric cones II: Beta-Riesz distribution

Generalized Univariate Distributions and a New Asymmetric Laplace Model

Log-Determinant Divergences Revisited: Alpha-Beta and Gamma Log-Det Divergences

Informative Bayesian inference for the skew-normal distribution

A Geometric View of Conjugate Priors

The Statistical Analysis of Series-Parallel System Products with Geometric Distribution under the Full Sample