Extensions of Fisher Information and Stam's Inequality
Authors:
- 27 February 2012
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Theory
- Vol. 58 (3) , 1319-1327
Abstract
We explain how the classical notions of Fisher information of a random variable and Fisher information matrix of a random vector can be extended to a much broader setting. We also show that Stam's inequality for Fisher information and Shannon entropy, as well as the more generalized versions proved earlier by the authors, are all special cases of more general sharp inequalities satisfied by random vectors. The extremal random vectors, which we call generalized Gaussians, contain Gaussians as a limiting case but are noteworthy because they are heavy-tailed.Keywords
This publication has 16 references indexed in Scilit:
- On the geometry of generalized Gaussian distributionsJournal of Multivariate Analysis, 2009
- Moment-Entropy Inequalities for a Random VectorIEEE Transactions on Information Theory, 2007
- CramÉr–Rao and Moment-Entropy Inequalities for Renyi Entropy and Generalized Fisher InformationIEEE Transactions on Information Theory, 2005
- A mass-transportation approach to sharp Sobolev and Gagliardo–Nirenberg inequalitiesAdvances in Mathematics, 2004
- The Kotz-type distribution with applicationsStatistics, 2003
- A characterization of the multivariate distributions maximizing Renyi entropy2003
- Best constants for Gagliardo–Nirenberg inequalities and applications to nonlinear diffusionsJournal de Mathématiques Pures et Appliquées, 2002
- Information-theoretic inequalities for contoured probability distributionsIEEE Transactions on Information Theory, 2002
- Symmetric Multivariate and Related Distributions1990
- Differential-Geometrical Methods in Statistics1985