Higher order models for fuzzy random variables (Q834459)

The present paper connects previous studies about fuzzy random variables (frv) with the theory of imprecise probabilities. In classical frv-theory, precise probability values are assigned to the events. In this paper, the authors deeply investigate a model to represent the available imprecise information about the ``true'' probability distribution of a frv. They define an acceptability function over the class of all possible probability distributions which can be interpreted as possibility distribution. Thus, the information about the true probabilities is represented by a possibility measure over the class of all probability measures (so-called second order possibility distribution). The authors present three different types of second order possibility measures. The first one, the fuzzy probability envelope, is the most informative one. The authors observe that the classical probability measure of a frv does not determine the fuzzy probability envelope, i.e. the probability measure induced by the frv is not the most suitable way to describe this type of imprecise information.