A new method to analyze evidence conflict
一种新的证据冲突分析方法

Dempster Shafer evidence theory is widely used in data fusion. However, incorrect results may be obtained by classical Dempster combination rule when collected evidence highly conflict with each other. Extensive work has been done to improve the DS combination rule, but how to measure the conflict among two pieces of evidence is an issue which has not been well considered. We define a new factor, the relative coefficient, for measuring the conflict between two pieces of evidence based on the partial entropy and the mixture entropy. When the relative coefficient is close to one, the conflict degree between two pieces of evidence is very low. On the other hand, when the relative coefficient is close to zero, the conflict degree between two pieces of evidence is very high. Some numerical examples are given to illustrate the efficiency of the proposed method.