The aim of this paper is to survey and briefly discuss various rules conditioning proposed in the framework of possibility theory as well as various conditional independence relations suggested for these rules. These conditioning rules and conditional independence relations are confronted with formal properties of conditional independence. Special attention is payed to the conditioning rule based on measure-theoretical approach. It is argued that this way of conditioning and the related conditional independence notion not only generalize some of presented rules and conditional independence relations, but also their properties correspond to those possessed by stochastic conditional independence.
Keywords. Possibility measure, possibility distribution, conditioning rule, natural extension, conditional possibility distribution, possibilistic conditional independence, formal properties of conditional independence.
The paper is available in the following formats:
Laboratory of Intelligent Systems
Related Web Sites