Uncertainty management in rule based systems application to maneuvers recognition

Abstract

In this paper we study the uncertainty management in expert systems. This task is very important especially when noised data are used such as in the CASSICE project, which will be presented later. Indeed, the use of a classical expert system in this case will generally produce mediocre results. We investigate the uncertainty management using the Dempster-Shafer theory of evidence and we discuss benefits and disadvantages of this approach by comparing the obtained results with those obtained by an expert system based on the fuzzy logic.