| Rough set theory, introduced by Z. Pawlak in the early 1980s of the 20th century, is a powerful mathematical theory of reasoning about data. Since 1990s of the last century, it has attracted much attention of researchers around the world and become a focus in the fields of computer science and information science. With more than twenty years development, rough set theory has been successfully applied to many areas including machine learning, pattern recognition, decision analysis, process control, knowledge discovery from databases, and expert systems.Enclosing the methods based on rough set in inconsistent information system some problems are discussed in this paper and some significative results obtained.As for attribute reduction in decision table, p lower approximation reduction based on variable precision rough set is analyzed aiming at difficulties in application of relative attribute reduction on rough set model. The concept of distribution reduction is introduced to variable precision rough set mode and β lower approximation distribution reduction is proposed. In addition, based on information entropy measure, the definition of attribute significance measure in decision tables based on variable precision rough set is given. Weak consistent information and strong consistent information in threshold of β are discerned. Uncertainty measure of consistent information outside positive region in standard rough set is depicted. The heuristic algorithm of β lower approximation distribution reduction based on information entropy is proposed. The reduction keeping the same decision classification is obtained in meaning of variable precision. These results provide theoretical basis and applied foundation for attribute reduction in inconsistent information system. The feasibility of algorithm is indicated using a concrete instance.As for decision rules obtaining, A. Skowron default rules obtaining methods is analyzed aiming at the problem of decision rules obtaining in...
|
PDF Full Text Request