Information Measures: the Curious Case of the Binary Alphabet

Abstract

Four problems related to information divergence measures defined on finite alphabets are considered. In three of the cases we consider, we illustrate a contrast which arises between the binary-alphabet and larger-alphabet settings. This is surprising in some instances, since characterizations for the larger-alphabet settings do not generalize their binary-alphabet counterparts. Specifically, we show that $f$-divergences are not the unique decomposable divergences on binary alphabets that satisfy the data processing inequality, thereby clarifying claims that have previously appeared in the literature. We also show that KL divergence is the unique Bregman divergence which is also an $f$-divergence for any alphabet size. We show that KL divergence is the unique Bregman divergence which is invariant to statistically sufficient transformations of the data, even when non-decomposable divergences are considered. Like some of the problems we consider, this result holds only when the alphabet size is at least three.

to appear in IEEE Transactions on Information Theory

Cited by in corpus (11)

• An elementary introduction to information geometry
• On a generalization of the Jensen-Shannon divergence
• Justification of Logarithmic Loss via the Benefit of Side Information
• Bounded rational decision-making from elementary computations that reduce uncertainty
• On $w$-mixtures: Finite convex combinations of prescribed component distributions
• Divergence and Sufficiency for Convex Optimization
• Maximum Entropy and Sufficiency
• Bregman Divergence Bounds and Universality Properties of the Logarithmic Loss
• The Bregman chord divergence
• Dependence Measures Bounding the Exploration Bias for General Measurements
• The space of positive transition measures on a Markov chain