Efficient composing rough approximations for distributed data

摘要

With the development of network and sensor technologies, distributed information processing has been involved in many scientific research fields. Rough set theory has been proved to be a powerful tool to deal with uncertain, inconsistent and fuzzy information. In rough set methodologies, computing rough approximations is a key and time-consuming step for attribute reduction and rule extraction. In this paper, we introduce a matrix-based rough set approach for processing data distributed across different sites. The approach can efficiently compose rough approximations of global concepts in the distributed information system by utilizing the existing rough approximations of local concepts in subsystems. Some numerical examples are employed to verify the feasibility of the approach. The corresponding algorithm is developed and the experimental estimations show that our algorithm outperforms its counterpart on the computational time.