TY - JOUR KW - Boolean Algebra KW - Continuous DSmT KW - Evidence Theory KW - Probability AU - Frédéric Dambreville AB -

When implementing the DSmT, a difficulty may arise from the possible huge dimension of hyperpower sets, which are indeed free structures. However, it is possible to reduce the dimension of these structures by involving logical constraints. In this paper, the logical constraints will be related to a predefined order over the logical propositions. The use of such orders and their resulting logical constraints will ensure a great reduction of the model complexity. Such results will be applied to the definition of continuous DSm models. In particular, a simplified description of the continuous impreciseness is considered, based on impreciseness intervals of the sensors. From this viewpoint, it is possible to manage the contradictions between continuous sensors in a DSmT manner, while the complexity of the model stays handlable.

BT - Information & Security: An International Journal DA - 2006 DO - http://dx.doi.org/10.11610/isij.2004 LA - eng N2 -

When implementing the DSmT, a difficulty may arise from the possible huge dimension of hyperpower sets, which are indeed free structures. However, it is possible to reduce the dimension of these structures by involving logical constraints. In this paper, the logical constraints will be related to a predefined order over the logical propositions. The use of such orders and their resulting logical constraints will ensure a great reduction of the model complexity. Such results will be applied to the definition of continuous DSm models. In particular, a simplified description of the continuous impreciseness is considered, based on impreciseness intervals of the sensors. From this viewpoint, it is possible to manage the contradictions between continuous sensors in a DSmT manner, while the complexity of the model stays handlable.

PY - 2006 SP - 85 EP - 103 T2 - Information & Security: An International Journal TI - Ordered DSmT and its Application to the Definition of Continuous DSm Models VL - 20 ER -