Concentration functions and entropy bounds for discrete log-concave distributions

Abstract

Two-sided bounds are explored for concentration functions and R\'enyi entropies in the class of discrete log-concave probability distributions. They are used to derive certain variants of the entropy power inequalities.