This paper investigates the computation of posterior upper expectations induced by imprecise probabilities, with emphasis on the consequences of Walley's concepts of irrelevance and independence. Algorithms that simultaneously handle imprecise priors and imprecise likelihoods are derived through linear fractional programming; sequences of independent measurements are then analyzed, and a result on the limiting divergence of posterior upper probabilities is presented. Algorithms that handle irrelevance and independence relations in multivariate models are analyzed through graphical representations, inspired by the popular Bayesian network model.
Keywords. Convex sets of probability measures,linear and linear fractional programming,graphical models and directed acyclic graphs.
The paper is available in the following formats and sites:
Av. Prof. Mello Moraes, 2231
|Fabio Gagliardi Cozmanfirstname.lastname@example.org|
Related Web Sites
Informal Introduction to Quasi-Bayesian Theory (and Lower Probability, Lower Expectations, Choquet Capacities, Robust Bayesian Methods, etc...) for AI
Software to compute posterior upper expectations in Matlab
The JavaBayes system