Maximum entropy, fluctuations and priors

Abstract

The method of maximum entropy (ME) is extended to address the following problem: Once one accepts that the ME distribution is to be preferred over all others, the question is to what extent are distributions with lower entropy supposed to be ruled out. Two applications are given. The first is to the theory of thermodynamic fluctuations. The formulation is exact, covariant under changes of coordinates, and allows fluctuations of both the extensive and the conjugate intensive variables. The second application is to the construction of an objective prior for Bayesian inference. The prior obtained by following the ME method to its inevitable conclusion turns out to be a special case of what are currently known under the name of entropic priors.

presented at MaxEnt 2000, the 20th International Workshop on Bayesian Inference and Maximum Entropy Methods (July 8-13, Gif-sur-Yvette, France))

References in corpus (1)

• Change, time and information geometry

Cited by in corpus (14)

• Maximum Entropy and Bayesian Data Analysis: Entropic Priors
• Is Tsallis thermodynamics nonextensive?
• Entropic Inference
• Change, time and information geometry
• Pairwise maximum-entropy models and their Glauber dynamics: bimodality, bistability, non-ergodicity problems, and their elimination via inhibition
• Using Relative Entropy to Find Optimal Approximations: an Application to Simple Fluids
• Entropic Priors
• A Bayesian Framework for Collaborative Multi-Source Signal Detection
• Maximum entropy approach to the theory of simple fluids
• Maximum Entropy approach to a Mean Field Theory for Fluids
• Maximum Entropy and the Variational Method in Statistical Mechanics: an Application to Simple Fluids
• Local quantum information dynamics
• Bayesian Entropic Inverse Theory Approach to Implied Option Pricing with Noisy Data
• First measurement of the small-scale spatial variability of the rain drop size distribution: Results from a crucial experiment and maximum entropy modeling