Research On Variable Precision Rough Sets And Fuzzy Similarity Measure

In the field of data mining, fuzzy set theory and the classical rough set theory hasimportant theoretical and practical value.This paper describes the development of roughsets and the research significance of fuzzy mathematics, Second, we introduce thebasics of rough sets and fuzzy similar theoretical;According to the attribute reduction ofrough set model, combined variable accuracy rough set attributereduction with dominance relations attribute reduction. Based on dominance relation,this paper constructed upper approximation and lower approximation. Usingupper(lower) distributed coordination set construction identification attributeset and matrix, proposed dominance reduction based on variable accuracy. This methodmakes up the attribute reduction vacancy on rough set theory. Finally this text mainpoint out that inside and outside accumulate approach degree which has been supportedby current approach degree axiom system emerging abnormal phenomenon, and analyzethe reason for the abnormal phenomenon. Then we should raise new axiom system todescribe the similarity of fuzzy vector from two direction. We also verify that Hammingapproach degree, Euclidean algorithm, maximum algorithm, arithmetic averagealgorithm, absolute value index method, absolute value counting backwardtechnique,absolute value subtraction and dot product fit Distance approach degreeaxiom. Correlation coefficient method,Cosine method fit inner product approach degreeaxiom. Then we structure the approach degree that combined by distance and inclinedangle.