Weighted Evidence Combination With Ranking Distance

The theory of belief functions, also called Dempster-Shafer evidence theory, is a powerful tool for uncertainty modeling and reasoning. However, one might encounter counter-intuitive results when using Dempster's rule in evidence combination. To handle this problem, an alternative weighted combination method is proposed in this paper, where the weights are generated according to the ranking distance between those evidences to combine. Experimental results show that our proposed approach can obtain rational combination results when compared with prevailing methods.