Title :
Combination of Non Distinct Evidences
Author_Institution :
Laboratory for Medical Statistics, School of Public Health, Brussels Free University (ULB)
Abstract :
The impact of an evidence can be descibed by a belief function. When 2 evidences are obtained, they can be combined with Dempster´s rule of combination if the evidences are distinct. We define here the concept of distinctness and develop the rule of combination that applies when evidences are non distinct.
Keywords :
Computational modeling; Laboratories; Power measurement; Public healthcare; Q measurement; Statistics;
Conference_Titel :
American Control Conference, 1984