A new correlation coefficient of BPA based on generalized information quality

Abstract

Dempster rule is commonly used to combine evidence from different sensors because of its excellent mathematical properties (commutativity and associativity). However, the conflict coefficient k of this rule cannot be reasonably and effectively express the conflict between evidence, which makes the Dempster rule greatly questioned. To describe in the relationship between evidence (conflict or correlation) more accurately, an evidence correlation coefficient based on generalized information quality is proposed. First of all, according to Deng entropy to measure the uncertainty of each evidence, and combined with the information quality of Yager, new generalized information quality is proposed, which performs well in measuring the basic probability assignment's certainty. Second, the evidence itself is modified by generalized information quality, and the evidence correlation coefficient is calculated based on the Pearson coefficient formula. A new measurement method of evidence conflict based on evidence correlation coefficient is proposed. Finally, combined with the evidence correlation coefficient and DEMATEL model, the evidence is discounted and combined. Numerous examples are used to analyze and compare the evidence conflict coefficient and the evidence combination results. The experimental results demonstrate that, compared with other evidence conflict measurement methods, the evidence conflict coefficient calculated by this method can reflect the difference between evidence more effectively. The result of evidence combination is more reasonable and accurate.

A new belief entropy and its application in software risk analysis

“…A is called the focal element when m(A) > 0. m(A) indicates the belief value that supports the proposition A. There are some studies about mass function [9,44,47], which has been used in many fields [2,24].…”

A New Total Uncertainty Measure from A Perspective of Maximum Entropy Requirement

“…It is widely used in fault diagnosis [11][12][13][14], decision-making [15,16], risk assessment [17], and so on. Many studies in recent years have focused on conflict resolution [18][19][20], evidence revision [21], combination rules [22][23][24][25], and information volume [26,27]. Many methods about uncertainty quantification have also been proposed [28].…”