Rough Concept Lattice-based Multi-attribute Decision Analysis

Rough set theory and Concept Lattice theory are two different mathematical methods for represent the uncertain knowledge. RS theory was proposed by Z. Pawlak in 1982. The focus of RS theory is on the ambiguity caused by limited discemibility of objects in domain of discourse. Concept Lattice theory was proposed by Wille in 1982 and based on data on objects and attributes and the establishment of relations between the dual concept of a level of theory, vivid succinctly embodies the concept of the generalized and specialized relations. The combination of concept lattice and rough sets are a new study and is very value in fact.Concept lattice is an effective tool for data analysis and knowledge discovery, with the accuracy and completeness, and other characteristics. Is a kind of rough set theory to deal with uncertainty, inaccurate data, the mathematical tools. Although the rough set theory and concept lattice theory research information systems are in the information uncertainty, but both the starting point, and so on, the characteristics are different, they are complementary. At present, the rough set theory and concept lattice theory combining the study of literature rarely, using the rough set theory in the approximate upper and lower sets, especially the concept of characterization and description of the extension of the uncertainty, and its rules of access to decision-making methods were studied, so that concept lattice said that the uncertainty of the knowledge capacity, it improved the ability of deal with uncertainty knowledge, had important theoretical and practical significance. The main research work are as follows:First, a new concept lattice structure:rough concept lattice RCL, and its constructor. Using rough set, similar to the next set, the definition of the concept of a cell has two connotations of the extension, that is similar to the extension and under similar extension. For a given cell of the connotation of rough concept lattice, which has a approximation of the extension with the nature of the uncertainty, and under similar extension is accurate, this reflects the structure and characteristics of the object and determine the uncertainty between the two kinds of relationships. The second discusses the rough concept lattice of the rules of access to decision-making methods and attribute reduction. Finally, the use of decision-making matrix to identify the background and gives the decision-making function to identify the background attribute reduction of the specific attribute reduction methods.