A Research Of Multi-granularity Rough Set Model Of Incomplete Fuzzy Order Information System

With the rapid development of Internet technology,data processing has been paid more and more attention by people.By analyzing data,we can get a lot of useful information for us.For example: analyzing the data of investment and wealth management can provide us with a risk reference before we decide to invest.Analysis of the user's purchase and browsing behavior data can provide businesses with the basis for accurate promotion of products.And many of these data are in order,such as student achievement assessment,investment risk rating assessment,thesis quality assessment.Rough set technology has its unique advantages in processing data.It does not need to provide any prior information other than data,so it has become an effective tool for processing data.Introducing the idea of granular computing into rough sets points out new data processing direction.In life,due to various reasons,the data obtained is often incomplete,with lossing or missing data.This paper considers the absence of data,attribute ambiguity,and the existence of sequentiality.The method of rough set and granular computing is used to study the corresponding model.The specific work is as follows:Firstly,we have improved the traditional relationship of extended dominance relation and proposed a strict definition of extended dominance relation.Based on the strict relationship,a new multi-granular rough set model is established.The traditional dominance relation is a kind of non-strict dominance relation.The article first introduced the concept of strict dominance relation,obtained an improved extended dominance relation,and discussed some of its related properties.Further from the perspective of optimism and pessimism,the multi-granularity rough set model is discussed and the corresponding definitions are given.Since the importance of each attribute may be different in real life,considering the weight of each attribute,a weighted multi-granularity rough set based on an improved extended dominance relation is proposed.Secondly,to improve the traditional restriction dominance relation,a multi-granularity rough set model based on the improved constraint dominance relation is proposed.The traditional restriction dominance relation is a kind of non-strict dominance relation.Therefore,the concept of strict dominance relation is introduced,and the improved restriction dominance relation is obtained.Because the extended dominance relation is too loose when dealing with problems and the error rate is high when the data is missing,the restriction dominance relation is introduced.The multi-granularity rough set model is discussed from two perspectives of pessimism and optimism and corresponding definitions are given.Finally,considering that the importance of each attribute is sometimes not the same,a weighted multi-granularity rough set model based on the improved restriction dominance relation is proposed.Thirdly,the existing rough set model of incomplete fuzzy sequence information system is extended.When the data lacks a lot,combined with the extended dominance relation,a multi-granularity rough set model based on the extended dominance relation is proposed to consider both optimistic and pessimistic.In this case,the corresponding model definition is given.When there is a large number of missing data,the concept of restricted dominance relation is adopted,and a multi-granularity rough set model based on restricted dominance relation is proposed.Finally,the proposed model's granularity reduction,rule extraction and attribute reduction are discussed,and some properties that they satisfy are proved.