Update of approximations in composite information systems

In this paper, a new composite rough set model is proposed to process incomplete composite information with criteria and regular attributes. Following that, some strategies for incrementally updating approximations of composite rough sets when adding or removing some objects are discussed and the corresponding incremental algorithms are designed. Several numerical examples illustrate the feasibilities of the composite rough set model and the incremental strategies, respectively. Experimental evaluation shows that the incremental algorithms can effectively reduce the running time than their counterparts. 2015 Elsevier B.V. All rights reserved.