A New PCR Combination Rule for Dynamic Frame Fusion

Abstract

-Dynamic frame fusion which is based on hybrid DSm model is an important problem in information fusion. But the traditional combination rules are mainly under fixed discernment frame (Shafer model and free DSm model) responding to static model. A new method for dynamic proportional conflict redistribution rules (dynamic PCR rules) based on hybrid DSm model is proposed for the shortness of classical dynamic PCR rules. In the new dynamic PCR rule, combination involved with empty set is defined as one kind to obtain more reasonable results. For the redistribution weight, the conjunction Basic belief assignment (BBA) and conflict redistribution BBA are both taken into account to raise the fusion precision. The effectiveness of revised dynamic PCR rule is studied and simulated in both aspect of fusion accuracy and calculation.

A Data Classifier Based on Maximum Likelihood Evidential Reasoning Rule

“…DST defnes a "frame of discernment" (FoD) including all class labels and uses BPA called "evidence" to measure the belief degrees that a sample with multiple attributes (features) belongs to every class and the subsets of these classes. Ten, some classical evidence combination mechanisms (Dempster's rule [21], evidential reasoning rule [21], Yager's rule [22], Smets's rule [23], DSm rule [24], Dubois and Prade's rule [25], etc.) can be used to fuse the multiple pieces of evidence so as to reducing the uncertainty of multifeature classifcation.…”