Research On Multi-granulation Rough Set Based On Set-valued Information System

Rough Set Theory is a powerful mathematical tool which can be used to deal with incomplete and uncertain information. At present, this theory has been successfully applied to various fields such as data mining. With the deepening of the study, the classical rough set model has been unable to satisfy the increasingly complex practical problems. As a generalized model of single-valued information system, set-valued information system can be used to characterize uncertain and insufficient information, and can also handle a lot of uncertainty flexibility. As an important branch in information processing, granular computing is playing a more important role in soft computing, knowledge discovery and data mining. Therefore, the study of multi-granulation rough set model based on set-valued information system, not only enriches and improves the classical rough set theory, but also is expected enormous application value due to the widespread applied background of these theories.Some theoretical problems were mainly researched in the paper as follows, and we got the following innovative results:1) The promation model of rough set based on set-valued information system is studied, the advantages of studying rough set by set-valued information system are analyzed, and the theoretical study of granular computing and rough set theory is given. The classical multi-granulation rough set model and variable multi-granulation rough set model are discussed, and their basic properties are analyzed as well.2) The multi-granulation rough set model based on set-valued information system is proposed. Also, by further exploring the basic properties and measure relations of this new model, the acquisition of decision rules is presented. This new model has the following innovations:First, because of the unity of set-valued information system, the model has a general representation in granular rough set theory; second, the model can make an effective judgment of audit risk on security auditing of information systems; finally, it can be seen from examples that the model can effectively weaken the authority influence among decision-makers, so that the decision-making process and the results have more objectivity and fairness.