French | English

Main Members Research Documentation Seminars at LRTS Photos Publications Links Events Management

|All|Fields|Specific Projects|Authors|Thesis| Application du calcul d'incidence à la fusion de données Generalized Incidence Calculus is a hybrid symbolic-numeric approach to data fusion that presents many interesting characteristics, in particular a correspondence with the Theory of Evidence. This master's thesis presents modifications to Generalized Incidence Calculus for its application to eliminate the Data Looping problem which makes combination of correlated data take more importance. Those modifications also allow the representation of alternative combinations of the Theory of Evidence by using a possible worlds approach. In particular, it is possible to associatively combine data using the Yager combination and parallels can be made with the Dezert-Smarandache Theory. par Marc-André Dumas déposé le 21 juillet 2006 sous la direction de Dominic Grenier dans le département Génie électrique