To some extent, uncertainty exists in most of real systems. Therefore, it is very important how to deal with these kinds of information, and intelligent information proceeding is getting more and more concerned by scholars and becomes a very hot point in theory and application of information science. People need computer to automatically recognize and manage indefinite phenomena in impersonal world. like human cerebra.Fuzzy set theory, put forward by American cybernetics expert Zadeh, is mathematician tool to conduct the inaccurate phenomena. It focuses on the blurring of sets, and explains the extent of inaccuracy of sets using the concept of membership function.Rough set theory, put forward by professor Pawlak, in the early of 20 century 80's, is a method to study the expression and leaning of incomplete or uncertain knowledge base. The method focuses on the roughness of sets, uses no-division concept to express the roughness of sets.Naturally a combination of these two theories can be searched since fuzzy set theory and rough set theory can be used to express knowledge through observing and testing datum. That is, the fuzzy rough set theory comes into being. When the knowledge described in approximate unite is fuzzy notion. Under such circumstance we may analyze rough approximation problems by rough set theory, and study roughness of fuzzy rough set by fuzzy set theory.In this article, after recalling the basis concepts of fuzzy set and rough set , anew definition of fuzzy rough set is offered, and the deference between the definition and that of given in document is discussed, the approximate accuracy is larger than that of given in document. Using fuzzy set theory, fuzzy rough conversion is defined and its properties are discussed, and the extension theorem of fuzzy rough set is given. Based on fuzzy rough set theory and probability theory, a kind of probability fuzzy rough set is given first, moreover, it's properties are given and its application in medical diagnose is discussed. |