Rough set based related affect attribute reduction algorithm

Abstract

Data mining is hot-spot research areas recently. Based on rough set theory, this paper proposes the definition of attribute real-time and its attributes reduction algorithm. By configuring the related affect matrix of the attributes and arousing with the attributes affects, we are able to delete the redundant attributes and get the reduction set which can reflect the mutual affect among the attributes. By using decision tables, the proposed algorithm can form the decision rules and serve the automatic diagnosis. The experiments show that our algorithm is more efficient than previous methods. © 2009 IEEE.