Two maximization problems of Rényi entropy rate are investigated: the maximization over all stochastic processes whose marginals satisfy a linear constraint, and the Burg-like maximization over all stochastic processes whose autocovariance function begins with some given values. The solutions are related to the solutions to the analogous maximization problems of Shannon entropy rate.

This paper introduces the variational Rényi bound (VR) that extends traditional variational inference to Rényi’s α-divergences. This new family of variational methods unifies a number of existing approaches, and enables a smooth interpolation from the evidence lower-bound to the log (marginal) likelihood that is controlled by the value of α that parametrises the divergence. The reparameterizati...

A remarkable yet mysterious property of black holes is that their entropy is proportional to the horizon area. This area law inspired the holographic principle, which was later realized concretely in gauge-gravity duality. In this context, entanglement entropy is given by the area of a minimal surface in a dual spacetime. However, discussions of area laws have been constrained to entanglement e...

In this paper, we propose a generalization of Rényi divergence, and then we investigate its induced geometry. This generalization is given in terms of a φ-function, the same function that is used in the definition of non-parametric φ-families. The properties of φ-functions proved to be crucial in the generalization of Rényi divergence. Assuming appropriate conditions, we verify that the general...

Csiszár’s forward -cutoff rate (given a fixed 0) for a discrete source is defined as the smallest number such that for every , there exists a sequence of fixed-length codes of rate with probability of error asymptotically vanishing as . For a discrete memoryless source (DMS), the forward -cutoff rate is shown by Csiszár [6] to be equal to the source Rényi entropy. An analogous concept of revers...

We show that a recent definition of relative Rényi entropy is monotone under completely positive, trace preserving maps. This proves a recent conjecture of Müller–Lennert et al. Recently, Müller–Lennert et al. [12] and Wilde et al. [15] modified the traditional notion of relative Rényi entropy and showed that their new definition has several desirable properties of a relative entropy. One of th...

A time-frequency distribution provides many advantages in the analysis of multicomponent nonstationary signals. The simultaneous signal representation with respect to the time and frequency axis defines the signal amplitude, frequency, bandwidth, and the number of components at each time moment. The Rényi entropy, applied to a time-frequency distribution, is shown to be a valuable indicator of ...

The redundancy for universal lossless compression in Campbell’s setting is characterized as a minimax Rényi divergence, which is shown to be equal to the maximal α-mutual information via a generalized redundancy-capacity theorem. Special attention is placed on the analysis of the asymptotics of minimax Rényi divergence, which is determined up to a term vanishing in blocklength.