Granule Description Based On Possible Attribute Analysis And Its Comparative Analysis

Granular computing proposed by Professor Zadeh is an effective method to simulate human thinking and solve large-scale complex problems.One of the important contents of granular computing is information granulation.Information granulation can be simply understood as the process of obtaining a granule layer under a given granulation criterion,which is a necessary process of granular computing.In the process of using the idea of granular computing to solve complex problems,the complex problems will be divided into multiple sub-problems according to certain granulation criteria,and then sub-problems will be described and dealt with.Therefore,obtaining reasonable granule descriptions through appropriate granulation criteria can make decision makers have a more thorough understanding of granules,which is helpful to solve complex problems.Granule description is the basis of using granular computing to solve complex problems,and it is also a very important and valuable research direction in granular computing theory.At present,the research direction of granule description is mainly based on common attribute analysis and necessary attribute analysis.The so-called common attribute analysis refers to the description of the object granule with the attributes common to all objects in the object granule,while necessary attribute analysis refers to the attributes common to all objects in the granule,it also involves the attributes owned by some objects,and these attributes will not be owned by other objects out of the granule.Thus it can be seen that common attribute analysis and necessary attribute analysis emphasize the common attribute and the unique attribute of the granule respectively,but these two analysis methods are too strict in solving practical problems,which will cause it not to make realistic decisions in team building,task allocation and other issues considering collective cooperation.Therefore,inspired by common attribute analysis and necessary attribute analysis,this paper proposes possible attribute analysis,which is used to study the granule description of definable and undefinable granules in a complete formal context.The details are as follows:1.Introducing the idea of stability index,stability index of attribute-oriented concept is defined,and the minimal generator is defined by the stability index,then the positive and negative definable granules are described in a complete context.2.The upper and lower approximate description of the undefinable granules are carried out by using the description of the definable granules,and the description error of the approximate description is defined,and then the optimal approximate description of undefinable granules is given from the view of description error.3.Discuss the relationship between possible attribute analysis and common attribute analysis,necessary attribute analysis,and use specific examples to illustrate the application value of possible attribute analysis proposed in this paper in solving specific problems such as team building,task allocation and so on.