The Study Of Attribute Reduction In Rough Set Theory

Rough set theory is a new tool deal with fuzzy and uncertain knowledge which is proposed by Pawlak in the early 1980s. It is a mathematical theory used to data analyse. The main idea is to approximate inexact, uncertain concepts by using of known knowledge of information. Now, this theory has become a flash point in the research area of international science. But it is proved to be a NP-Hard problem in seek the smallest reduction of any information system. So it is the emphases of reduction algorithm research to seek heuristic information and contruct heuristic algorithm.The equivalence relations play important roles in classical rough set theory. Started with the equivalence relations of classical rough set model, to the problems among attribute reduction, Such as, attribute reduction is very sensitive to dates of noises. So it's very significant to depict the extract relationship and uncertain relationship among attributes under the condition of noises interference. Rough set theory and information theory are combined in this paper with advantages are exerted and disadvantages are limited between each other. A new attribute algorithm is presented. On the basis of research of previous attribute reduction algorithms. A modified heuristic algorithm of attribute reduction is presented in this paper. This algorithm adopts an idea of divide and rule. Two improved definitions of attribute significance are proposed considering the influence of information area of distributing to knowledge compartmentalization. The shortage of predefined weight of condition attribute in calculating weighted sum attribute significance is pointed out and the weight is modified at the same time. With a view of characteristics of analysissitus, corresponding heuristic information is constructed. Finally, the validity and feasibility of the algorithm is demonstrated and applied in gyration analyse of the biggest shipping.