Information Structure,Uncertainty Measurement And Attribute Reduction Of Two Kinds Of Incomplete Information Systems

Information system is an important model in the field of artificial intelligence.Incomplete information system is an information system with missing value,incomplete set-value information system and incomplete interval-value information system are two kinds of incomplete information systems.This paper mainly studies the information structure,uncertainty measure and attribute reduction of these two kinds of incomplete information systems.Their information structure,uncertainty measure and attribute reduction have important research significance.Chapter 1,the research background,research status and main results of this paper are briefly introduced.Chapter 2,tolerance relations on object sets are constructed,their information structures are proposed by defining the similarity between two information values of each attribute in two incomplete information systems,dependency between information structure is discussed,and some properties of information structures are given.Chapter 3,information granulation,information entropy,rough entropy,information amount and traditional methods are proposed to measure the uncertainty of two kinds of incomplete information systems.Based on this,the optimal choice of information structure in two kinds of incomplete information systems is obtained.In addition,the uncertainty measurement method of rough sets in two kinds of incomplete information systems is also proposed.Chapter 4,?-reduction is proposed in two incomplete information systems,the relation between the proposed ?-reduction and uncertainty measurement is established,and then attribute reduction algorithms based on ?-discernable matrix,?-information granulation,?-information entropy and ?-importance degree is proposed.