Title :
A Set of Three New Conflict Measurement Methods for Different Evidence Condition
Author :
Jing Zhu ; Chen-Xi Wang ; Li-Fang Hu ; Zi-Ye Long
Author_Institution :
Dept. of Automatio, Tsinghua Univ., Beijing, China
Abstract :
In the combination of conflict evidences under belief function theory framework, the first step is to reasonably measure the conflict between evidences. Firstly, shortcomings and limitations of the existing conflict measurement methods are analyzed. Secondly, general criterion of conflict measurement between evidences is put forward. Then a set of three new conflict measurement methods for different evidence conditions is designed. The new methods can not only satisfy the interchangeability and combinability, but also overcome the operational problem of the existing binary conflict measurement method. Compared with other methods, the new methods are more suitable for people´s intuitive logic reasoning, and have the best comprehensive effect under different evidence condition.
Keywords :
belief networks; case-based reasoning; belief function theory framework; binary conflict measurement method; combinability; conflict evidences; conflict measurement methods; evidence conditions; interchangeability; intuitive logic reasoning; Automation; Cognition; Complexity theory; Distance measurement; Finite element analysis; Time measurement; Vectors; basic belief assignment; basic belief mass; conflict evidence; information fusion;
Conference_Titel :
Measuring Technology and Mechatronics Automation (ICMTMA), 2013 Fifth International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4673-5652-7
DOI :
10.1109/ICMTMA.2013.33