Entropy and divergence rates for Markov Chains. II. The weighted case

Proceedings of the Romanian Academy Series A-Mathematics Physics Technical Sciences Information Science

Vlad Stefan BARBU¹ 

关键词: divergence measures;    information measures;    Markov chains;    entropy;    divergence rates;    weighted divergence measures.;   

DOI  :  

学科分类：计算机科学（综合）

来源: Editura Academiei Romane

【 摘 要 】

In this work we consider weighted versions of the generalizations of Alpha divergencemeasure (Chernoff [8], Amari and Nagaoka [1]) and Beta divergence measures (Basu et al. [4]) forMarkov chains and investigate their limiting behavior. This is a continuation of the developmentspresented in [2], applying the notion of weighted divergence measure (Beliş and Guiaşu [5], Guiaşu[11], Kapur [12]). This work is continued in [3], where we present generalized Cressie and Read powerdivergence class of measures and numerically investigate some properties of all these generalizeddivergence measures and rates.

【 授权许可】

Unknown