On Pinsker's Type Inequalities and Csiszar's f-divergences. Part I: Second and Fourth-Order Inequalities
Gustavo L. Gilardoni
Abstract
We study conditions on f under which an f-divergence Df will satisfy Df ≥ cf V2 or Df ≥ c2,f V2 + c4,f V4, where V denotes variational distance and the coefficients cf, c2,f and c4,f are best possible. As a consequence, we obtain lower bounds in terms of V for many well known distance and divergence measures. For instance, let D(α) (P,Q) = [α(α-1)]-1 [∫ qα p1-α d μ-1] and Iα(P,Q) = (α-1)-1 [∫ pαq1-α d μ] be respectively the relative information of type (1-α) and Rényi's information gain of order α. We show that D(α) ≥ 1/2 V2 + 1/72 (α+1)(2-α) V4 whenever -1 ≤ α≤ 2, α= 0,1 and that Iα = α2 V2 + 1/36 α(1 + 5 α- 5 α2) V4 for 0 < α< 1. Pinsker's inequality D ≥ 1/2 V2 and its extension D ≥ 1/2 V2 + 1/36 V4 are special cases of each one of these.
Create a lesson
Related papers
Deterministic Identification over Additive Gaussian Channels
Jonathan E. W. Huffmann, Holger Boche
Knowledge Distillation Driven Semantic NOMA with GAN Refinement for 6G Robotic Vehicle Networks
Qifei Wang, Zhen Gao, Li Qiao et al.
Real-Time Reconstruction of Markov Sources over MPR Channels
Pansee S. Elessawy, Nikolaos Pappas
Minimum Rate For Partially Observable Linear System with Side Information: LQG Plant and Gaussian-Markov Source
Sijie Li, Hyeji Kim
Spectral Approximation and Ergodic-Capacity Convergence of HMIMO Channels under Spatial-Wavenumber Domain Mismatch
Hangsong Yan, Hong Yang, Shu Sun
Sharp Minimax Regret for Infinite-Memory Logistic Prediction
Vaneet Aggarwal