Extraction Method of Classification Rules in Decision Tree Based on Attribute Selection Metric of Rough Set

摘要

In this paper, the kernel algorithm of decision tree is used as a tool, and on the basis of constructing a good data source group for classification, the characteristics of decision tree classification algorithm are analyzed and experimented. This paper introduces a rough set attribute selection metric, which selects attributes for classification from the point of view of improving the accuracy of classification and the purity of sub-databases. Association rule mining is the mining of interesting association rules or relationships between indexes (items) in the database. The paper present extraction method of classification rules in decision tree based on attribute selection metric of rough set.