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.

Some Relationships and Properties of the Hypergeometric Distribution

Another look at binomial and related distributions exceeding values close to their centre

On Some Properties of the Beta Normal Distribution

Two Monotonicity Results for Beta Distribution Functions

On Beta-Product Convolutions

Bayesian Inference for the Gamma Zero-Truncated Poisson Distribution with an Application to Real Data

An Infinite Product of the Incomplete Beta Function-type Hypergeometric Function and its Probabilistic Origins

Two Generalizations of the Binomial Distribution

The generalized Poisson-binomial distribution and the computation of its distribution function

An Algorithm for Computing the Distribution Function of the Generalized Poisson-Binomial Distribution

Hypergeometric Distribution Revisited: Tail Inequalities, Confidence Bounds and Sample Sizes

Two-dimensional generalized gamma function and its applications

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

Conjugate Analysis Of The Conway-Maxwell-Poisson Distribution (Vol 1, Pg 363, 2006)

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

On the gamma difference distribution

Bounds for novel extended beta and hypergeometric functions and related results

A generalized inverse gausssian distribution with τ-confluent hypergeometric function.

Some Characterizations of the Extended Beta and Gamma Functions: Properties and Applications

Bacon With Your Eggs? Applications of a New Bivariate Beta-Binomial Distribution