Yet Another Paradigm Illustrating Evidence Fusion (YAPIEF)

In information fusion situations, it is vital to manage uncertainty and the incomplete information set for time critical decisions. The Dempster-Shafer evidential theory is a very elegant method of mathematically representing this knowledge. The evidential theory knowledge is represented by a frame of discernment with a power-set number of basic belief assignment (BBA) components. For real time implementation this may be a bit of a conundrum especially when supporting many hypotheses in real time systems. A multi/dual probability delineation of Dempster-Shafer evidential theory is presented to overcome the power-set problem for real time implementations. The set of BBAs are mapped via pignistic probability transforms to many sets of probabilities that support the incomplete information set. These sets of probabilities are ordered via the probability information content equation for further processing (fusing, decision making). The problem has been transformed from addressing a power-set of components 2^Omega to a multiple number of probability components N Omega greatly simplifying real time implementations. A fusion process demonstrating dispersion in decision space is also presented