Strong Asymptotic Composition Theorems for Mutual Information Measures

Benjamin Wu, Aaron B. Wagner, Ibrahim Issa, G. Edward Suh

Abstract

We characterize the growth of the Sibson and Arimoto mutual informations and $\alpha $ -maximal leakage, of any order that is at least unity, between a random variable and a growing set of noisy, conditionally independent and identically-distributed observations of the random variable. Each of these measures increases exponentially fast to a limit that is order- and measure-dependent, with an exponent that is order- and measure-independent.
BibTeX
@article{wu2024strong,
  author = {Benjamin Wu and Aaron B. Wagner and Ibrahim Issa and G. Edward Suh},
  title = {{Strong Asymptotic Composition Theorems for Mutual Information Measures}},
  journal = {IEEE Trans. Information Theory},
  year = {2024},
  url = {https://openreview.net/forum?id=ERvZ0RmutE},
}

Loading related content...

Figure