A Method of Knowledge Acquisition Based on the Discernibility in Imperfect Information System

摘要

Traditionally speaking, the information system is assumed to be perfect, i.e. attribute values are not missing and supposed to be precise. In fact, imperfect information system (include missing data and imprecise data) is always exist. However researches on the method of knowledge acquisition based on imperfect information system are relatively less. In this paper, based on imperfect information system, a method of attribute reduction on the basis of the discernibility of information system is proposed. The example shows that the method can find the reduct of imperfect information table and the relative reduct of imperfect decision table. These results have important instructive significance for further studying on the knowledge acquisition in imperfect information system.