A compromise operator-based approach for multigranulation space.

The notion of compromise operators was systematically explored for multi-attribute decision making problems, which include widely used averaging operators, uninorms and nullnorms as the special cases. In this paper, we employ compromise operators to examine the issue of information fusion in multigranulation spaces. For this purpose, we firstly show that the optimistic multigranuation rough set model can be interpreted from the viewpoint of uninorm operators and the pessimistic multigranulation rough set model can be interpreted from the viewpoint of averaging operators or nullnorm operators. Then, by considering rough membership degrees in each Pawlak space and using the generalized compromise operator, we present a novel approach to information fusion in multigranulation space. Lastly, an illustrative example of information fusion in multigranulation spaces is presented.