Evidential weights of multiple preferences for competitiveness

The alternative selections or feasible sets in multi-attribute utility theory usually face problems of inaccuracy or inconsistency in weighting. Evidential weight and weight of evidence still have not resolved these problems in an efficient manner. This research proposes evidential weights of multiple preferences by using evidence and inference based on the rough set theory. An evidential probability, is constructed to resolve this weight. Technically, the evidential probability is required to satisfy the following conditions: uniqueness, independence and consistency, to identify various evidence and noise, and to provide quantities of certain evidence to resolve weights. Empirically, the resolved weights are fused into competitiveness utility for verification. A practical case successfully classifies dominating nations of World Competitiveness Yearbook. Another case analyzes the competitiveness trends of China during 1997 to 2012, predicting China will continue rising in near future.