BeliefNet Tool: An Evidential Network Toolbox for Matlab

Walid Trabelsi and Boutheina Ben Yaghlane.

The BeliefNet Tool is a new engine to perform local computations efficiently and conveniently in both undirected and directed evidential networks (i.e. networks using belief functions theory [10], [12]). BeliefNet Tool is written in Matlab and it implements the basic algorithms as well as the main operations performing inference of beliefs in evidential networks. Attention is especially devoted to present the structures used when implementing BeliefNet Tool.

PDF full paper