| Nowadays,human society is having great progress,and the information processing technology is progressing.Information systems are characterized by multiple forms,large scale and complex structure.Especially,the problem of dealing with fuzzy object information is particularly prominent.In order to improve the accuracy and objectivity of the description of the fuzzy object,the concept of intuitionistic fuzzy set is proposed by K,T.Atanassov.It includes not only the membership degree of fuzzy objects,but also two parts of non membership degree and hesitation degree,therefore,the information of the intuitionistic fuzzy set contains the fuzzy object more complete and more specific.The theory of intuitionistic fuzzy sets is another tool to deal with the uncertainty and fuzzy problem of data after Pawlak's rough set theory is put forward.Since the intuitionistic fuzzy set theory is introduced,the research and application of intuitionistic fuzzy sets have attracted the attention of scholars at home and abroad.And its research achievements have been widely applied to many fields,such as financial risk analysis,information rule extraction,data mining and market investment prediction.In this paper,under the background of intuitionistic fuzzy information system,rough membership and rough set model of probability decision theory are studied.Under the intuitionistic fuzzy order information system,the distributed attribute reduction model,the assignment attribute reduction model and the relative knowledge granularity attribute reduction model are studied.At last,the rough set model of decision theory after fusion is studied under the background of multi source information system.The main innovation points are as follows:1.In the intuitionistic fuzzy information system,the fuzzy equivalence relation is constructed,and the fuzzy approximate space is formed.In the fuzzy approximation space,the rough membership degree is redefined,and its related properties are proved.An example is given to verify the validity of the related properties and further expand the scope of uncertainty measurement.2.By using the membership degree and non membership degree of the intuitionistic fuzzy set,the intuitionistic fuzzy order relation is constructed.And an intuitionistic fuzzy order decision information system is established.Three attribute reduction models are constructed in the intuitionistic fuzzy order decision information system:(1)the distribution attribute reduction model is established by using the distribution function and the distribution identification matrix;(2)the assignment attribute reduction model is established by using the assignment function and the assignment identification matrix;(3)the attribute reduction model is constructed by using the relative knowledge granularity and the internal and external importance of the attributes.The theory of attribute reduction is further enriched.3.Based on the intuitionistic fuzzy information system,using the theory of rough sets,Bias decision theory and probability measure,a probabilistic decision analysis model based on intuitionistic fuzzy information system is established.Based on fuzzy approximate space and intuitionistic fuzzy approximation space,the probabilistic decision rough set theory method of intuitionistic fuzzy information system is studied.Thus,a more accurate decision is made for the decision-makers.4.Based on multi-source information system,three kinds of multi-source fusion decision rough set method are put forward: conditional entropy multi-source decision fusion(CE-MSD),decision support degree multi-source decision fusion(DS-MSD)and mean multi-source decision fusion(M-MSD).Three corresponding algorithms are designed to verify the effectiveness and feasibility of the proposed method.Finally,a better multi source fusion decision rough set method is obtained by the experiment of the UCI data set. |
PDF Full Text Request