Improved measure of evidence conflict based on pignistic probability distance
Huan Liu, Yilin Fang, Quan Liu, Aiming Liu · 2016
Dempster-Shafer Theory has been extensively employed in the field of information fusion due to its impressive capability to cope with uncertain information. But when it deals with completely conflictive evidences or highly conflictive evidences, the phenomenon of counter intuition will occur. In this paper, shortcomings of the commonly accepted measure of conflict are analyzed and to avoid the wrong results made by using the conventional method, pignistic probability function is introduced to discriminate the differences between the evidences. Furthermore, based on the introduced pignistic probability function, an improved method for measuring the degree of conflict is proposed and proved. And then, the evidences are combined via a new combination rule to yield the ultimate result. The illustrative examples certify that even when the evidences are highly conflictive, targets can be recognized effectively.