A Lower Bound for the Fisher Information Measure
Information Theory
2015-06-16 v5 math.IT
Abstract
The problem how to approximately determine the absolute value of the Fisher information measure for a general parametric probabilistic system is considered. Having available the first and second moment of the system output in a parametric form, it is shown that the information measure can be bounded from below through a replacement of the original system by a Gaussian system with equivalent moments. The presented technique is applied to a system of practical importance and the potential quality of the bound is demonstrated.
Keywords
Cite
@article{arxiv.1306.2607,
title = {A Lower Bound for the Fisher Information Measure},
author = {Manuel Stein and Amine Mezghani and Josef A. Nossek},
journal= {arXiv preprint arXiv:1306.2607},
year = {2015}
}
Comments
4 pages, submitted to IEEE Signal Processing Letters