Generalised exponential families and associated entropy functions
arXiv:0803.0104 · doi:10.3390/entropy-e10030131
Abstract
A generalised notion of exponential families is introduced. It is based on the variational principle, borrowed from statistical physics. It is shown that inequivalent generalised entropy functions lead to distinct generalised exponential families. The well-known result that the inequality of Cramer and Rao becomes an equality in the case of an exponential family can be generalised. However, this requires the introduction of escort probabilities.
20 pages
Cited by in corpus (23)
- Group entropies, correlation laws and zeta functions
- Escort Evolutionary Game Theory
- Uniqueness and characterization theorems for generalized entropies
- Some properties of generalized Fisher information in the context of nonextensive thermostatistics
- Nonlinear statistical coupling
- Unifying Aspects of Generalized Calculus
- Optimal measures and Markov transition kernels
- Quantum Statistical Manifolds
- A new entropy based on a group-theoretical structure
- Theory versus experiment for vacuum Rabi oscillations in lossy cavities
- On a (β,q)-generalized Fisher information and inequalities involving q-Gaussian distributions
- On the thermodynamics of classical micro-canonical systems
- Maximum entropy estimation of transition probabilities of reversible Markov chains
- The q-exponential family in statistical physics
- Notions of the ergodic hierarchy for curved statistical manifolds
- Thermodynamics of exponential Kolmogorov-Nagumo averages
- On multidimensional generalized Cramér-Rao inequalities, uncertainty relations and characterizations of generalized -Gaussian distributions
- Non-extensivity of the configurational density distribution in the classical microcanonical ensemble
- Data Set Models and Exponential Families in Statistical Physics and Beyond
- Entropies from coarse-graining: convex polytopes vs. ellipsoids
- Bistable stochastic processes in the q-exponential family
- On the connection between linear combination of entropies and linear combination of extremizing distributions
- Two types of -Gaussian distributions used to study the diffusion in a finite region