Incremental updates of Dominance-based approximations

DRSA (Dominance-based Rough Set Approach) is an extension of rough set theory for dealing with multiple criteria decision analysis problems based on dominance principle. In this paper, we consider dominance-based approximation spaces based on a generalized dominance principle which are represented by indexed blocks of binary neighborhood systems. We introduce algorithms for updating approximation spaces of decision classes when decision examples are added to and removed from a multiple criteria decision table incrementally. The proposed algorithms are demonstrated by examples.