基于概念格同构下的属性约简及其算法研究

ON ATTRIBUTE REDUCTION AND ITS ALGORITHM BASED ON CONCEPT LATTICE ISOMORPHISM

通过研究概念格和同构理论,发现不同的概念格之间存在同构关系,并引入算子E=A-D和G=E-D**,给出一系列概念格同构的判定定理。基于概念格同构这一理论,对形式背景的属性约简及最小属性约简集的算法进行初步研究,为应用概念格进行数据挖掘提供新的理论依据。

Through the study of concept lattice and isomorphism theory,we find that between different concept lattices there are the isomorphic relationships,and introduce the operator E = A- D and G = E- D＊＊,present a series of discriminant theorem of concept lattice isomorphism; Based on this concept lattice isomorphism theory,we make initial study on the attribute reduction and the minimal attribute reduction set algorithm of formal context,which provide a new theoretical basis for applying the concept lattice to data mining.