On Hölder projective divergences

Abstract

We describe a framework to build distances by measuring the tightness of inequalities, and introduce the notion of proper statistical divergences and improper pseudo-divergences. We then consider the Hölder ordinary and reverse inequalities, and present two novel classes of Hölder divergences and pseudo-divergences that both encapsulate the special case of the Cauchy-Schwarz divergence. We report closed-form formulas for those statistical dissimilarities when considering distributions belonging to the same exponential family provided that the natural parameter space is a cone (e.g., multivariate Gaussians), or affine (e.g., categorical distributions). Those new classes of Hölder distances are invariant to rescaling, and thus do not require distributions to be normalized. Finally, we show how to compute statistical Hölder centroids with respect to those divergences, and carry out center-based clustering toy experiments on a set of Gaussian distributions that demonstrate empirically that symmetrized Hölder divergences outperform the symmetric Cauchy-Schwarz divergence.

25 pages

Cited by in corpus (6)

• An elementary introduction to information geometry
• On the Kullback-Leibler divergence between discrete normal distributions
• Clustering in Hilbert simplex geometry
• Monte Carlo Information Geometry: The dually flat case
• A note on Onicescu's informational energy and correlation coefficient in exponential families
• Generalized Fisher-Darmois-Koopman-Pitman Theorem and Rao-Blackwell Type Estimators for Power-Law Distributions