On a posterior information process for parametric families of experiments
In a filtered statistical experiment a priori and a posteriori probability measures are defined on an abstract parametric space. The information in the posterior, given the prior, is defined by the usual Kullback-Leibler formula. Certain properties of this quantity is investigated in the context of so-called arithmetic and geometric measures and arithmetic and geometric processes. Interesting multiplicative decompositions are presented that involve Hellinger processes indexed both by prior and by posterior distributions.
|General theory of processes (msc 60G07), Applications of stochastic analysis (to PDE, etc.) (msc 60H30), Theory of statistical experiments (msc 62B15), Measures of information, entropy (msc 94A17)|
|Logistics (theme 3), Energy (theme 4)|
|CWI. Probability, Networks and Algorithms [PNA]|
Dzhaparidze, K.O, Spreij, P.J.C, & Valkeila, E. (1998). On a posterior information process for parametric families of experiments. CWI. Probability, Networks and Algorithms [PNA]. CWI.