Dynamic directed evidential networks with conditional belief functions: application to system reliability
Résumé
The temporal dimension is a very important aspect which must be taken into account when reasoning under uncertainty. The main purpose of this paper is to address this problem by a new evidential framework for modeling temporal changes in data. This method, allowing to model uncertainty and to manage time varying information thanks to the evidence theory, offers an alternative framework for dynamic probabilistic and dynamic possibilistic networks. It is applied to a system reliability analysis for the sake of illustration.