Information Geometry and Classical Cramér-Rao Type Inequalities
This work provides a theoretical extension for statisticians and information theorists, but it is incremental as it builds on existing geometric frameworks.
The paper tackles the generalization of classical Cramér-Rao type inequalities using information geometry, showing that this framework can be applied to derive four specific inequalities from different divergence functions, such as the α-version and Bayesian versions.
We examine the role of information geometry in the context of classical Cramér-Rao (CR) type inequalities. In particular, we focus on Eguchi's theory of obtaining dualistic geometric structures from a divergence function and then applying Amari-Nagoaka's theory to obtain a CR type inequality. The classical deterministic CR inequality is derived from Kullback-Leibler (KL)-divergence. We show that this framework could be generalized to other CR type inequalities through four examples: $α$-version of CR inequality, generalized CR inequality, Bayesian CR inequality, and Bayesian $α$-CR inequality. These are obtained from, respectively, $I_α$-divergence (or relative $α$-entropy), generalized Csiszár divergence, Bayesian KL divergence, and Bayesian $I_α$-divergence.