Bayesian approach to thermostatistics (Q1103316): Difference between revisions

Bayes' theorem is used to derive the dual of the Gibbs formulation of statistical thermodynamics. An asymptotic analysis is performed, akin to Khinchin's use of the central limit theorem, to determine approximate expressions for the moment-generating functions. The prior densities, which are determined by equating the maximum-likelihood estimates with the moment expressions in the asymptotic limit, satisfy Jeffrey's invariant properties of improper prior densities.