HUGIN architecture for propagating belief functions
In this paper we study the use of a HUGIN like architecture for propagating Dempster-Shafer belief functions. The main issue of the proposed scheme is that it allows to do the entire propagation with mass function. To achieve this, an inverse for mass functions is defined.
Main Authors: | , |
---|---|
Format: | info:eu-repo/semantics/report |
Language: | English |
Published: |
2012
|
Online Access: | http://hdl.handle.net/10835/1553 |