Evidential reasoning model based on extended attribute

In multiple attribute decision making problems, the variety of attribute becomes large because of the complex decision environment and uncertainties in the process of decision making. In the existent evidential reasoning approach, only benefit and cost attribute have been considered, so it could not deal with the decision problem with other kinds of attribute. In this paper, deviating attribute is added to the ER based MADM problems. Specifically, the representation of deviating attribute by evidence is analyzed, and the transformation rules from deviating attribute assessment values to belief degrees are studied. As the result, the applications of ER based assessment approach are enlarged.