Attribute Reduction Of Formal Decision Contexts Based On Three-way Concept Lattices

Formal concept analysis(FCA)which is a tool for data analysis and its foundation is formal context was put forward by German mathematician Wille.With developing of the research,FCA has been one of the most important topics in artificial intellegence,and been widely used in machine learning,picture processing and decision analysis.Attribute reduction whose basic idea is to find the minimal set of attribute to meet the given consistent is an important research direction in FCA,and the research about it mainly concentrate on classical concept lattices.The three-way concept analysis(3WCA)combines the FCA and three-way decisions(3WD),the object(attribute)induced three-way concept lattice have more information.In this paper,we study the method of attribute reduction in formal decision context based on three-way concept lattices.The contents are as follows:1.Based on the two kinds of three-way concept lattices,we study the attribute reduction in AE(OE)-three-way consistent formal decision context,which preserves the information of AE(OE)-irredundant rules without loss.2.We present the calculating approache which preserves the information of AE(OE)-irredundant rules without loss to attribute reduction.3.Based on the two kinds of three-way concept lattices,we define several types of attribute reductions in formal decision context and study their relationships.