A New Hybrid Associative Classification Algorithm Based on OR-Tree and Pruning Skills

摘要

A new hybrid associative classification algorithm based on Order-tree and pruning skills is proposed to solve the problems that there are always too much time consuming and the searching spaces are always too large in traditional associative classification algorithm. In the method a structure of the orderly rules of the tree (OR-tree) is been designed to store the information of the classification rules so that the association rules can be organized when generated, meanwhile, a series of pruning skills that face the rule sets and the candidate classification rules are proposed to optimize the OR-tree. The experiment results show that the new hybrid associative classification algorithm based on OR-tree and pruning skill can not only reduce the searching space of frequent sets and the time consuming effectively, but also improve the mining efficiency of the associative rules.